Mars is near the limit of where human beings can go on a small spacecraft. In a sense, it’s bad luck that Nature tempted us with a barely livable planet close enough to reach with rudimentary technology. If not for Mars, the next step for human spaceflight would be finding ways to build spaceships the size of ocean liners instead of trying to fly people around in a space Prius, and then we could go explore more interesting destinations in the Solar System.

But you go to space with the planets you’re dealt.

Traditionally, the chief strategies for dealing with radiation risk in Mars planning have been optimism and denial. Optimists posit the emergence of magic technologies (mostly drugs), while denialists argue that radiation levels on Mars are just a healthy part of clean Martian living.

While the radiation environment around Mars is now well-characterized, the biological effects of this radiation (which is difficult to simulate on Earth) remain a matter of conjecture. By current estimates1, a Mars-bound crew on a 1,000 day mission would face between a 2% and 15% chance of radiation-induced death during or after the mission. The uncertainty in the figure is almost all due to our deficient understanding of the non-targeted effects of heavy ion radiation, particularly at the chronic dose rate that Mars-bound astronauts would experience.

This uncertainty makes the long-running debate between ‘it’s not a problem’ and ‘it’s a showstopper’ factions impossible to adjudicate until we have better data.

Rather than stake out a side, I want to briefly sketch out the radiation problem opportunity and frame it in terms of tradeoffs.

Chronic vs. Acute Hazards

There are two different radiation hazards facing a Mars mission, both linked to the eleven year solar cycle. You can think of one as acute and the other as chronic.

High Energy Protons

The acute radiation hazard to a Mars mission comes from the Sun who, like Yosemite Sam, is forever firing off bursts of high-energy protons in all directions. The frequency of these bursts, called solar particle events (SPEs), rises and ebbs with the eleven year solar cycle. When the sun is at its most active, there may be several solar eruptions a day, while during solar minima there may be just one event every few days. Events can last from hours to days, and energetic ones will have an onset time measured in minutes.

These energetic burst of protons follow magnetic field lines outward from the Sun at a large fraction of the speed of light, and can carry enough energy to kill an unprotected crew. NASA was lucky to dodge one storm in August of 1972 that happened to fall between Apollo missions; if astronauts had been on the Moon at the time, they would have experienced acute and incapacitating radiation sickness.2

A Mars-bound spacecraft can expect to be hit by a half dozen or so proton events of unknown duration and severity. The spacecraft that carried Curiosity to Mars (at around the midpoint of a solar cycle) recorded five solar proton events during its eight month transit, and this seems like a reasonable baseline to work from for a human mission.

There are enough atoms in a spacecraft to give the crew significant protection from high-energy protons, provided they have advance warning. Most spacecraft designs include a shelter where the crew can go hide behind water tanks, propellant, or two years’ worth of their accumulated frozen feces, reducing their exposure by an order of magnitude during the peak hours of such an event. Modern designs also have a kind of plastic body armor for astronauts to wear over sensitive organs.

The biggest risk is being caught unprepared, particularly if any crew members are working outside the spacecraft when the burst arrives. No space suit can adequately protect against high-energy protons. Our ability to give a crew early warning depends on our planet’s position relative to the spacecraft. On a long-stay mission trajectory, where Earth usually sits along the lines of magnetic flux that connect the Sun and the spacecraft, the warning time can be as little as 22 minutes. On a short-stay profile, where the crew is on the far side of the Sun from Earth, there might be no warning at all.

The thin Martian atmosphere is an effective shield against most solar protons (those with energies below about 150 MeV). Astronauts on Mars can also build themselves a shelter out of rocks or ice to hide in. So it’s fair to consider solar protons a threat to crews only while they are in transit.

Galactic Cosmic Rays

The chronic radiation hazard on a Mars mission comes from little bits of charged supernova shrapnel that zip around the solar system in all directions, a crossfire of high-energy particles collectively called Galactic Cosmic Radiation (GCR).

The most vigorous of these particles can carry an impressive amount of energy. One celebrity cosmic ray in 1991 arrived on Earth at 99.999999999999999999999% the speed of light, a single proton packing the momentum of a little leaguer’s fastball. While such monster ions are rare, there are enough high-energy particles in the sky to make galactic cosmic radiation the biggest health risk to a crew. Unnervingly, astronauts can observe this radiation directly by closing their eyes and waiting for a particle to smash into their retina. These flashes are sometimes bright enough to wake them up from sleep, a personal memento mori arriving from a faraway star.

There’s no practical way to shield astronauts against the high energy component of galactic radiation. In fact, attempts at partial shielding make the problem worse, because high energy cosmic rays colliding with the metal walls of the spacecraft will create a cascade of highly penetrating secondary neutrons, increasing the total dose absorbed by the crew. (Something similar happens in the Earth’s atmosphere. There’s an area around 20km altitude called the Regener-Pfotzer maximum where the radiation dose from cosmic rays is much higher than at higher altitudes).

You can think of this as the Hürtgen Forest problem. If someone is firing artillery shells at you, instinct tells you to go hide behind something. But if you hide inside a grove of trees, the exploding artillery shells will send giant splinters of wood raining down on your foxhole, making you less safe than you were in the open.

The same situation obtains in space. The first few centimeters of shielding (the walls of the spacecraft) block a significant fraction of cosmic rays. But after that, things get worse before they get better. To thoroughly shield against cosmic rays would take a layer of water or plastic many meters thick,3 which is simply impossible until we’re willing to build a bigger rocket.

The best defense against cosmic rays is to stay home under a thick blanket of air and a good magnetic field. The second-best defense is to go fast and time one’s travel to coincide with the peak of the solar cycle, when the sun’s swelling magnetic field reduces the cosmic ray flux in the inner Solar System by about half. Unfortunately, that’s also the time the Sun will be spitting protons at you like an angry cat.

But since the risk from solar proton events and galactic cosmic radiation is anti-correlated, you have to pick your poison. And as terrifying as it sounds to fly during solar maximum, it’s better than absorbing the cumulative dose from a sky full of invisible particle accelerators taking pot shots right at you.

Some relative numbers

NASA measures radiation hazard in Sieverts, a derived unit that tries to convert the objective dose measured by physical detectors into a quantifiable measure of cancer risk.

There are problems with the Sievert! There are also problems with NASA’s career limit of 600 mSv! But I will ignore all of this and take the numbers at face value to get a relative sense of scale.

A typical ISS astronaut gets hit by 0.3 mSv/day of cosmic rays.4 An astronaut on the Moon, outside of Earth’s magnetosphere, would see a cosmic ray flux about four times stronger. On Phobos, the larger of the Martian moons, that flux would be a little higher still, since GCR intensity goes up with distance from the Sun. On Mars itself, where the atmosphere provides some degree of protection, the relative flux goes down to about 2.5 times what we see on the space station.

Finally there is free space, the long coast phase between the Earth and Mars. During this phase, when the only thing between the crew and shrapnel from every ancient supernova in the Local Group is a wall of boxed dinners, radiation exposure will be highest. Astronauts en route to Mars experience a cosmic ray flux about seven times higher than on the International Space Station.

So from a cosmic ray standpoint, every day spent in space is equivalent to about three days on Mars, or a week on the space station.

What does this mean for your Martian commute?

Recall that there are two broad classes of Mars mission, a long-stay profile and a short-stay profile. One is shorter but spends almost all of its time in space, the other is a year longer and includes a long residency on the Martian surface.

We can use these mission templates to get a rough sense of cosmic ray exposure. The diagram below shows a 1120 day long-stay mission that starts with a 90 day shakedown period in high Earth orbit. The red line shows the cumulative GCR dose at solar minimum; the blue line shows the same at solar maximum:

As you can see, the solar cycle is the biggest factor in radiation exposure. Flying at solar maximum exposes the crew to just over the NASA career limit, while flying at solar minimum doubles the dose. The daily dose astronauts receive on Mars is much lower than in space, but because of the length of the mission, it is a significant contributor to the total.

Here’s a similar chart for the short-stay mission profile:

Surprisingly, despite the shorter mission time (650 days vs. 1000) and very different trajectory, the total cosmic ray dose on both types of mission is roughly the same.

There’s not a lot we can do to radiation-proof a short stay mission. Its duration is dictated by orbital mechanics (the relative movements of Mars and Earth), and as I explained above, there’s no good way to add shielding to chemical rockets.

But with the long stay mission, there are options! One way to get the dose down is for the crew to bury their shelter under regolith (Martian rocks and soil). If the crew spend the first month of their time on Mars shoveling two meters of rocks onto their home, we get a cumulative radiation graph that looks like this:

This is better, but not by a lot, since most of the total exposure comes from those six-month trips to Mars and back.

So a second idea is to make those transits faster, shortening the punishing periods of full exposure in deep space. With orbital refueling, significant handwaving, and oceans of money, it would be possible to get transit times down to 90 days. In that case, exposure would look like this:

This is a reduction of one third, at a pretty steep cost in space bulldozers and magic rockets. Flying at solar maximum, meanwhile, gets you about a 40% reduction for free.

So one tempting lesson is that we should always fly at the peak of the solar cycle, solar protons be damned.

The trouble is, we never know what kind of solar maximum we’re in for. Not only is the peak of a cycle hard to predict (the 11 year duration is an average), but its magnitude as measured by sunspot activity can vary by a factor of three. Moreover, the most intense maxima tend to be compressed in time, so that the peaks can only cover a portion of a three-year mission.

A final fillip is that we have no guarantee that the peak of a solar cycle will line up with energetically favorable launch windows for Mars, which recur on a ~15 year cycle.

The porkchop plot below shows launch windows to Mars for the period 2030-2060. Areas in blue and green show a combination of departure date and transit time that is achievable with chemical rockets, while the orange and pink areas are energetically haram.

You can see that a great opportunity for a fast transit exists in the 2035-36 period and also happens to coincide with a solar maximum. After that, though, the two cycles go out of phase. The predicted solar maximum in 2046 coincides with bad orbital juju that makes it hard to reach Mars, while the excellent launch windows in 2050-2052 happen right at the predicted solar minimum.

All this makes it very important to figure out just how much radiation an astronaut can take. We can get bend the exposure curve down with a lot of effort, but do we really have to?

I can think of at least five important tradeoffs that depend on the answer.

Solar Maximum vs. Other Times

So far in this post I’ve made solar maximum sound like the greatest thing since the wheel. But the solar proton risk on short stay missions is very real, and rises quadratically with approach to the Sun.5 It would be nice to avoid it!

Solar storms can also badly interfere with communications, and make it harder to explore far from the spacecraft. For that reason, it would be nice to know if we can get away with visiting Mars in the radiation “shoulder season”, or even during solar minima.

If it turns out missions are limited to the period around solar maximum, that ties Mars exploration to the 11 year solar cycle and makes many exploration paradigms (like the inbound crew relieving and potentially rescuing active crew) impossible. It also means there may be decades when we can’t go to Mars at all.

Women vs. Old Men

One delicate issue is that radiation is kind of sexist. A space dad can take more of a radiological beating than a woman in her thirties. NASA’s blanket 600 mSv limit is based on a 3% chance of radiation exposure-induced death (REID). But this is a risk that varies a lot with age and sex. A 55 year old male astronaut could spend 256 days in deep space at solar minimum before hitting the limit, while a 35 year old woman would hit it after 159 days. (A 55 year old woman would have 201 days).6

This ties in to some other sex differences in space flight. Older women are of course more susceptible to bone loss, but women in general are also less of a burden on life support systems, which would make an all-female Mars crew easier to design for.

NASA insists that having a unisex career limit is an ethical stand, but it’s possible that Mars missions will only be feasible for the least radiologically sensitive astronaut cohorts, and that this may cut across some politically uncomfortable lines.

Mission vs. Lifetime Risk

Measures like REID don’t distinguish between the chance that an astronaut develops aggressive leukemia halfway through a mission, or six years after landing. But in practical terms, this matters a lot. If everyone gets radiation-induced cataracts on Mars and can’t see, then that crew is going to die, even though a similar affliction back on Earth would be eminently survivable.

The question of prompt risk also drives aspects of mission design, like whether to send chemo drugs along on a Mars mission (and figure out how to keep them out of the life support loop).

At a minimum, we need to know if risks that astronauts are exposing themselves to are ones that might have a rapid onset. This extends beyond cancer to particularly poorly-understood areas like heavy-ion-induced cognitive deficits and cardiovascular damage.

Abort Scenarios

Another tradeoff is planning for contingencies. Let’s say the crew arrives at Mars and the heat shield on their lander looks broken. Is it riskier to attempt the landing or stay on the spacecraft? Or how about a toxic leak in the habitat halfway through the surface mission? Is it safer to try to clean it up and live in the residue, or abort to orbit?

We need to know the extent to which a purely orbital long-stay mission (1,000 days in deep space) is a death sentence to be able to make these kinds of decisions. Note that this drives not just operations, but mission design. If abort-to-orbit is survivable, then the orbiting transit vehicle should be made oversized and outfitted as a potential lifeboat. If the crew absolutely has to stay on Mars, then there will need to be more redundancy (and more downmass) at the landing site. These decisions have to be made in the early design phase.

Contamination

Shielding on Mars has significant consequences for contamination risk. If you bury a human habitat under soil, you are creating a large warm area (complete with meltwater) that becomes a potential habitat to terrestrial microbes, and giving them a royal highway into the subsurface. The same holds true if you build the human habitat inside a cave or lava tube, natural radiation shelters that are among the most promising habitats for finding relic life on Mars.

We need to know if we can get away with keeping astronauts in the open and minimizing ground disturbance, or if health and survival concerns require digging up the ground. That in turn will affect the choice of landing site.

The Usual Dreary Conclusion

Longtime readers will know what’s coming next. To gather enough data to make these tradeoffs, we need to perform long-duration animal tests outside the Earth’s magnetosphere, and eventually send some volunteers up there to get irradiated.

This necessary research program front-loads Mars exploration with years of ISS-like missions to nowhere. On the bright side, it gives NASA something to do while we wait for the 15 year Mars transit cycle and the 11 year solar cycle to sync back up again in our sunset years.

But the upshot is that we should just! send! robots!

Further reading

1. The lion of all things Mars-radiation-related is Francis Cucinotta. For an excellent summary of the radiation problem, see How Safe Is Safe Enough? Radiation Risk for a Human Mission to Mars. (2013) DOI 10.1371/journal.pone.0074988

2. Cucinotta also has strong words about NASA’s 600 mSv radiation limit. A proposed change to astronaut exposures limits is a giant leap backwards

   for radiation protection (2021) DOI: 10.1016/j.lssr.2021.07.005

3. A terrific and exhaustive overview of the radiation environment on a Mars mission: Radiation environment for future human exploration on the surface of Mars: the current understanding based on MSL/RAD dose measurements (2021) DOI 10.1007/s00159-021-00136-5

My purpose, and my belief, is that the bombs that killed and maimed at Hiroshima and Nagasaki shall one day open the skies to man.

—Freeman Dyson, A Space Traveler’s Manifesto, 1958

The nuclear pulse rocket is what you’d get if you hired a 12 year old to get you to Jupiter. It works by farting a continuous string of nuclear bombs (at the rate of about one per second) out its back end and riding the ensuing blast waves on a giant shock absorber, like a pogo stick. A series of hundreds or thousands of nuclear detonations accelerates the spacecraft to pretty much any speed you want, and when it’s time to slow down, you just turn around and start nuking in the forward direction.

Simple, easy, and fun!

The performance on this thing is sensational. Rocket engineers have always been stuck having to choose between thrust and efficiency. Chemical rockets that are powerful enough to get things off the ground (like Saturn V or Starship) are hopelessly inefficient, while the efficient ion motors we put on probes and satellites have only a few ounces of thrust. It’s like forever being forced to choose between an electric tricycle and a top fuel dragster, with no middle ground.

Like an El Camino rolling coal, nuclear pulse rockets occupy that missing middle. The energy density of nuclear fuel gives them incredible miles to the mushroom cloud, while thrust is only limited by how much the hammering the spaceship can take before shaking apart. Where the Apollo rockets had six stages and a mass ratio of about 540:1 (for every kilo of astronaut or spacecraft that landed back on Earth, you needed more than half a ton of fully-fueled rocket on the launch pad), a nuclear pulse rocket has a mass ratio closer to 1.5. It can take off from Earth, land 4,000 tons of scientists and equipment on Mars, and come back in one piece to refuel, as many times as you want.

That kind a mass budget is what Mars mission planners call ‘ample’. Consider that the International Space Station, the biggest object ever assembled in space, weighs 400 tons. Nuclear pulse propulsion means no more worrying about life support or radiation. You can stock the inside with all the oxygen and frozen steaks a crew can eat, encase the whole thing in radiation-blocking plastic, cap it with a glass-domed rotating casino for the view, and still have room for the thousands of fission bombs (dispensed like coke cans) you will need to detonate to get the thing moving. A crew on such a rocket would travel to Mars in comfort and style and arrive refreshed.

For that matter, they could travel to Saturn and arrive refreshed. An early 1958 design envisioned sending a crew of 20 to Enceladus and back within a span of three years, or about as long as it would take to fly astronauts to Mars and back on a conventional mission using chemical rockets. And they could do it in a fully reusable vehicle on a single launch from Earth.

In short, the nuclear pulse rocket solves all of the problems that plague chemical rockets, albeit at the cost of replacing them with much bigger, scarier problems.

Here are some other representative missions enabled by nuclear thunder:

• Soft-land 5,700 tons on the Moon (compare to 17 tons for Apollo)

• Land a 1300 ton payload (three times the mass of the International Space Station!) on Enceladus and return it to Earth on a 3 year round trip.

• Send a crew of 20 on a two year round trip to Callisto or Europa (with enough shielding to make Europa survivable)

• Send a crew of 50 on a 200 day round-trip to Mars, with 30 day surface stay

• Send 10,000 tons to medium Earth orbit.

Unlike every other kind of spacecraft, the only size constraint on a nuclear pulse rocket is that it can’t be too small. A practical nuclear pulse rocket—and just typing the adjective ‘practical’ here kind of sets my heart racing—weighs around 4,000 tons, about the size of a decemt apartment building. But you get much better performance if you build one the size of a cruise ship, or a city.

Like so many good ideas, nuclear pulse propulsion started with a Polish guy living in New Mexico. While working at Los Alamos in the 1940s, the mathematician Stanisław Ulam sketched out an idea for a spacecraft that could be accelerated by small nuclear explosions behind it. Freeman Dyson and Ted Taylor (the Rembrandt of American nuclear weapons design) later fleshed out the idea at General Atomic and got it modestly funded in the wake of the 1957 Sputnik panic.

Even in the 1950s, it was hard to get anyone with budget authority to stop screaming long enough to appreciate the benefits of the design. And so the budgets for Project Orion were always stingy; no one wanted to take the responsibility of actually making the rocket happen. A toy version of the design was built and tested (with conventional explosives) in 1959, and the project came within a whisker of having a proper atomic test before luck and funding ran out in 1964.

It would have worked great! Like with so many nuclear technologies, whether or not it worked was really low on the list of problems with it.

Principles of operation

The nuclear pulse rocket as envisioned in 1958 was a 4,000 ton behemoth that looked like a cartoon bullet, attached by a forest of shock absorbers to a broad, flat pusher plate.

The plate was a flat disk of metal that weighed about 1/6 as much as the ship proper. It’s easiest if you imagine it as a piston in a big two-stroke engine.

The full propulsion cycle looks like this:

1. A small (0.1-3 kiloton) nuclear bomb detonates about a hundred meters behind the pusher plate. The explosion vaporizes a disk of propellant (this can be anything from ice to metal to the crew’s own stored waste) and flings it at the plate at a velocity of many thousands of kilometers per second.

2. The plasma cloud hits the pusher plate, accelerating it with the force of a cannon shot. The intense heat of the collision ablates away a thin layer of oil on the pusher plate, protecting the main body of the plate from damage.

3. Giant airbags and shock absorbers turn the impulsive acceleration on the pusher plate into a longer pulse gentle enough (2-4 g) to be endurable by a human crew, and transmit it to the spacecraft.

4. A spritzer spritzes the plate with a fresh layer of oil.

5. The compressed shock absorbers rebound and return the pusher plate to its original position just as another bomb reaches the detonation point.

You can think of the process like accelerating a golf cart by hitting it with a sledgehammer. How rough the ride is depends on the size of the spacecraft, the size of each bomb, and how far the shock absorbers travel in each stroke.

Subscribe now

The bombs go off about once or twice a second, and every explosion adds about 20 mph to the vehicle’s speed. It takes 200 bombs to get Orion out of the atmosphere, and another 600 or so to put the spacecraft in a 300 mile circular Earth orbit.

Normally we think of nuclear bombs as vaporizing everything around them. The Orion concept works by minimizing the time the plate spends in thermal contact with nuclear debris. A round trip to Mars and back might involve 2,000 nuclear detonations, but the pusher plate will have spent a total of less than one second in contact with superhot plasma. The same technique that lets you carry a hot potato comfortably by tossing it hand to hand gets you to the outer planets.

What could go wrong?

It’s a Gatling gun for atomic bombs, what do you think can go wrong? But Orion has some fascinating ways to fail besides the obvious.

One challenge is how to handle a dud. In normal operation, the pusher plate rebounds against each successive explosion like a racket bouncing off of a tennis ball. If a bomb fails to go off, the rebounding piston will want to fly off the back of the rocket into space. So you need a mechanism to arrest its momentum, along with special procedures for firing a half-charge to restore it from a hyperextended state to a neutral position.

Another interesting failure mode is a fizzle—a scenario where the chemical explosives in the nuclear bomb go off, but fail to detonate the nuclear pit. Remarkably, this is far more dangerous to Orion than an atomic blast. The pusher plate is designed to absorb the impact from a uniform cloud of hypervelocity plasma, not sharp chunks of shrapnel. So a lot of design work has to concentrate on making sure the pusher plate isn’t dinged up too badly by flying debris from a conventional explosion.

The most complicated part of Orion is the mechanism for delivering bombs to the correct spot behind the pusher plate. The bombs are heavy (several hundred pounds) and have to arrive at a point a couple hundred meters behind the spacecraft with precise timing. You are basically building an accurate, low-velocity machine gun for nuclear shells.

Several mechanisms were considered: one was a series of angled tubes arranged around the rim of the spacecraft, which would only be exposed when the piston was maximally compressed at the top of the stroke. Another was a rocket launcher that would send little nukes curving around the edge of the pusher plate, detonating when they reached a pair of crossed radar beams. Since the core problem (reliably delivering identically sized containers from a rack) had already been solved in Coca Cola vending machines, Orion famously consulted with that company to help with aspects of this design.

But the biggest technical obstacles facing Orion had to do with computers, not nukes. One was navigation; a big reason the early design had a crew of 20-40 was that so the boys in the nose could get to work with graph paper and sextants when it came time to steer the spacecraft.

Another was making sure bombs with the right yield were fired in the right sequence during takeoff (a problem that became trivial with dial-a-yield fission weapons developed just a few years later).

And there were fundamental questions of design. How turbulence in the plasma wave would interact with the pusher plate was a question only atomic testing could answer in 1959, though it would be easy to simulate on computers today.

Drawbacks

There are some drawbacks to the nuclear bomb rocket.

For one, there’s the matter of launching any of this into space. Unless you start nuking Florida from ground level (and I’m not saying ‘no’), you need dozens of traditional rockets to lift this affront to God and Nature into orbit. Even then, its fission products will get caught in the Earth’s magnetic field and eventually spiral their way back home.

But probably the biggest drawback of nuclear pulse propulsion is the thousands of nuclear bombs.

It was not lost on potential funders that whoever sat in the pilot’s chair of this rocket would immediately command one of the largest nuclear arsenals on Earth. All it needed to turn Orion into a Pez dispenser for Armageddon was to redirect the bombs a little bit so they rained down on Earth instead of exploding against a pusher plate.

The list of launch hazards for Orion was also spectacular. A serious accident on the launch pad would cook off thousands of tons of high explosive and contaminate a huge area downrange of the spacecraft with plutonium. An out-of-control Orion zigzagging around the sky would also be a memorable sight, for anyone who was not instantly blinded by it (another awkward hazard). And how exactly do you handle range safety on a 4,000 ton vehicle full of atomic bombs?

Or consider the question of misfires. While the ship could be built to tolerate duds, each misfire meant that someone along the flight path was waking up to an almost-functional nuclear bomb in their backyard.

Some of these risks could be mitigated by launching Orion from remote Pacific islands. But the problem of fallout was pernicious. The 200 explosions that get Orion out of the atmosphere are the equivalent in fallout terms to a ten-megaton air burst. Back in the 1950s, when we were firing multi-megaton H-bombs every other Tuesday, this didn’t seem so bad. But we live in more delicate times, and the problem doesn’t go away once the vehicle reaches space. Any bombs that explode in Earth orbit create clouds of charged particles that eventually fall back down on Earth. Unlike the ascent plume, this fallout is not localized, and settles on the just and unjust alike.

Efforts to ameliorate the fallout problem run us into the final drawback, which is a little bit more abstract. To get the most out of its nuclear fuel and minimize fallout, Orion needs to use shaped nuclear charges with very low yield, that consume as little fissile material as possible. But these happen to be the worst kinds of nukes imaginable from a proliferation standpoint.

No one is really afraid of terrorists blowing up a big, meaty H-bomb that has to be carted around in a truck and needs a couple hundred kilograms of plutonium to detonate. But at the low end, nuclear weapons can be made frighteningly small and light. Ted Taylor describes a working six-inch atomic bomb that he could hold in one hand, while strongly hinting that bombs made with even smaller amounts of fissile material are feasible, provided you’re willing to wrap them in a lot of conventional explosive.

Unfortunately, addressing the fallout problem exacerbates the political problems that doomed Orion in the first place. The best rocket design we have unavoidably pushes the frontiers of nuclear miniaturization, a technology that is inherently destabilizing to a fragile status quo.

And that’s why no one today is riding Orion to the stars.

Benefits

Let’s not be negative Nellies, though. There is also a lot of good to say about the nightmare bomb rocket!

Engineers have been so beaten down by the rocket equation that it can be hard for them to grasp the full possibilities of a spacecraft unconstrained by weight. Rocket design has always been about squeezing the last drop of performance out of lightweight materials. Nuclear pulse rocket design is more about making sure the billiards table stays level during acceleration, and that the ship’s wine cellar strikes the right balance between tannic reds and dry whites. In his blueprints for Orion, Ted Taylor always made sure to include a heavy barber’s chair, just to drive home the point that this was a new class of vehicle.

Only three components on a nuclear pulse rocket need to be built to a high engineering tolerance: the shock absorbers, the bomb dispenser, and the bombs themselves. Everything else can be riveted together out of whatever material is handy. In other words, the nuclear pulse rocket is the thing Starship aspires to be: a rocket that is cheap to mass-produce, fully reusable, and capable of rapidly colonizing the solar system. The only hitch is that no one wants to give Elon nukes.

The fact that nuclear pulse rockets are huge and fast solves the hardest problem of interplanetary travel: protecting crews from radiation. Galactic cosmic radiation is annoyingly dangerous and extremely penetrating, to the point where you need something like ten meters of polyethylene or water to provide protection on par with Earth’s atmosphere. Galactic cosmic rays also have the awkward property that partial shielding can be worse than none, since secondary radiation created when high energy particles collide with the shielding can do more damage than the particles themselves.

Because GCR flux gets worse the further out you go in the solar system, and because voyages to the outer planets would necessarily last many years, cosmic rays set a practical limit to human exploration with chemical rockets. The moon is reachable, Mars is marginal, but anything further than that is off-limits.

There is no way a chemical or nuclear thermal rocket could ever carry adequate shielding against cosmic rays1. But Orion can. Nuclear pulse rockets can be built large enough and fast enough to make trips to Jupiter or Saturn with low cumulative radiation exposure, and armored with enough shielding to protect the crew not only from cosmic rays, but from the intense planetary radiation at desirable destinations like Europa.

And the bigger you make your rocket, the easier it gets to throw in artificial gravity, solving the other big physiological problem of space flight. As I wrote about in an earlier post, you need a structure about 112 meters in diameter to get a livable rate of spin at 1g. If you’re using chemical rockets, you’re limited by the fairly narrow diameter of the rocket fairing, and so spacecraft big enough to provide spin gravity have to be assembled in orbit. But 112 meters is a perfectly workable diameter for nuclear pulse rocket. All you have to do is spin one up once it’s headed in the right direction and live comfortably on the walls until you reach your destination.

As a last benefit, nuclear pulse rockets solve the entry and landing problem on distant worlds through brute force. Instead of worrying about aeroshells and supersonic retropropulsion, you can just null out your orbital velocity, then make judicious use of nukes to slow your descent straight downwards. If you’re squeamish about nuking your landing area, you can always switch to parachutes or chemical retro-rockets at the last minute, or deploy a dedicated lander, like one of those mega-yachts that carries a smaller superyacht as cargo.

The point is you’ll be landing 1950s torch-ship-style—slowly and under full control, without needing to husband every ounce of propellant. Abundance!

The Four Eras of Project Orion

We came very, very close to building this thing!

Like Taylor Swift, the nuclear pulse rocket went through several distinct eras in its development.

Read more

Share Mars For The Rest of Us

The first impression you get when visiting Biosphere 2 is that it would be a great place to start a cult. The futuristic greenhouse complex is built in the hills north of Tucson and has a beautiful view of the Santa Catalina mountains. The environment contains dangerously high levels of Arizona woo—you can practically feel the earth currents flowing down your energy meridians and making your various chakras and nadis vibrate in sympathy with Mother Gaia.

And the people who ran the place were kind of cultlike! They practiced mandatory weekly meditation and radical theater, were based out of a place called Synergia Ranch near Santa Fe, and had a charismatic and domineering leader named John Allen, who would harangue them for hours about world history and the noösphere. It didn’t help that the eight biospherians who locked themselves inside for a two-year stint in 1992 dressed in Star Trek-like uniforms and had names like Laser and Harlequin. There was a distinct Hale-Boppness to the whole enterprise that people at the time picked up on, years before the Heaven’s Gate cult finally gave wearing matching track suits a bad name. This was not a group of people whose mortal shells you would have been surprised to find arranged in a mandala on the floor when the doors finally re-opened on their great experiment in 1993.

But in truth, the biospherians were sane and competent, and they achieved something that has eluded everybody else, then and since. They built a large-scale habitat and lived inside it for two years in conditions of full material closure, growing their own food, recycling their own waste, and maintaining a breathable atmosphere in a completely sealed environment. Their large complex of greenhouses leaked less air than even the Space Shuttle1, and the 17 months they spent inside before needing to supplement the atmosphere with outside oxygen remains the absolute record for survival in an environmentally closed system.

Biosphere 2 was built as a complex of conjoined greenhouses, each hosting a different type of biome. The site included a miniature ocean, complete with wave machine, and a living coral reef. There was a savannah room, a tropical rain forest, a fog desert (an arid microclimate with high humidity) along with a farming area and the biospherians’ living quarters. The group shared a communal kitchen, but each resident had a private space he or she could retire to. Underneath the whole complex was a machine room, also part of the enclosure. Two large rooms with weighted bellows, called the lungs, equalized air pressure with the outside world and served as storage rooms for excess biomass (cut grass) when carbon dioxide levels got too high.

Today the complex is run as an ecological lab by the University of Arizona, and you are free to go inside on a tour. The whole facility feels a bit like a video game level, or a place where you might have to frantically search for the murderer while your fellow biospherians dropped one by one, with the police watching helplessly from outside.

The biospherians ate only what they could grow, a diet built around sweet potatoes, bananas, beets, goat milk, and a particularly vile-tasting bean called the lablab. They spent about three quarters of their time doing farm work, meal prep or cleanup, and approximately all of their time being hangry, which contributed to tensions that split the crew into two irreconcilable factions within a year of starting the experiment.

The meal schedule had crew members taking turns to cook lunch and dinner for their shipmates, which ensured that the two worst cooks on the crew each prepared a quarter of the meals. Their diet was mostly vegetarian, supplemented with small amounts of eggs, milk, and occasional meat and fish. Breakfast was invariably porridge sweetened with papaya, and any attempt at innovation around this meal (the only chance during the day to feel full) met with threats of violence. Lunch was a soup. There was a midmorning snack of peanuts, counted out carefully so everyone got the same portion; the hungry crew ate them with the shells on.

Like in any subsistence society, life in the biosphere became entirely centered on food. There was a regular schedule of feast days (American holidays and crew birthdays) to offset the constant gnawing hunger. The biospherians ate so many sweet potatoes their skin turned orange from the beta carotene.

Pest control was a battle. Stowaway Australian cockroaches covered the floors at night and destroyed everything they could get their mandibles on, including much of the salad crop and both of the kitchen’s microwave ovens (they ate the wiring). The broad mite, a pest that normally afflicts tea plants, broadened its horizons in this special setting and killed the entire potato crop. In the Ocean, the biospherians engaged in a battle of wits with stowaway octopi who had been accidentally introduced into the habitat and whose voracity was only eclipsed by their cunning. After the first few weeks they were never seen, but their depredations continued.

By objective criteria the crew were aggressively healthy—everyone lost weight, blood pressures and serum cholesterol dropped, and they lived the dream of a modern life extension influencer. But they were also grouchy, beset by cravings for sugar and coffee, and towards the end of the mission (as oxygen levels in the sealed habitat approached 14%) chronically hypoxic, to the point of sleep apnea. Early in the mission one crew member destroyed most of the coffee crop by damaging the flowers on the young coffee plants. In her book on the experiment, Jane Poynter recounts being so hungry that she stole a handful of monkey chow (it tasted sweet), along with multiple occasions when she and her boyfriend cooked themselves a illicit handful of sorghum late at night in the lab. The only locked room in the whole biosphere was the door to the banana pantry; no one trusted themselves to resist the sweet fruit at night.

The Biosphere missions were meant to keep going, but after the first experiment, the project’s billionaire benefactor got into conflict with management and pulled the plug. For a while the Biosphere was run by Steve Bannon(!), then got dumped on Columbia University, and finally landed in the hands of the University of Arizona, who operates it today. There are no more experiments with environmental closure, but the unique facility has enabled a lot of ecological research that would be difficult or impossible to do in a less controlled environment. You can go visit it today—get there early in the morning, and you’ll practically have the place to yourself.

Biosphere 2 was a bit of a punch line, but it deserves a better reputation than it got. This unlikely project remains our best window into what living in a Martian habitat would be like.

Ecology vs. Engineering

There are two conceptual approaches you can take to off-world farming. One is the NASA approach of treating agriculture as an engineering problem, where plants are just a specialized green tool in a mechanical toolkit. In this paradigm, you want to minimize the number of constituent parts needed to keep the crew alive, while excluding all extraneous factors. Your goal will be to cultivate hardy single-species plants that are free of pests, microorganisms, or any other complications, in known and reproducible conditions.

The term for this kind of minimalist farming is gnotobiology, from the Greek root gnosis. The idea is that everything in your biological system is fully characterized, down to the microbe level. At the extreme end, you might choose to grow organisms that are axenic, with no other species present at all. (This is particularly easy to do for fish, since the outside of their eggs can be sterilized before they hatch.) But since many plants need bacterial partners to thrive, it often describes systems where you’ve excluded everything except a short whitelist of chosen, beneficial organisms.

The other approach to farming is ecological. Instead of building up from a minimal set of species, you start from the other end, trying to capture as much of the natural community around an organically grown crop as you can fit in a sealed enclosure. This will necessarily include a lot of mystery organisms, including all kinds of insects, some pests, and a vast and unknown microbiome. The motivating idea is that you stand a better chance of preserving the balance that exists in a natural ecosystem on Earth by sizing it down than you would by trying to build it up one species at a time, from first principles.

Biosphere 2 incorporated both approaches. For habitats that were well-understood (savannah, desert, and of course the farm plot), biospherians carefully curated the set of species, from plants to insects, that they wanted to have in their world. But for the most complex habitats (ocean and rain forest) they used the other approach, collecting bulk soil and water samples from natural sources, and then trying to create conditions that would keep them thriving in captivity.

The advantage of the gnotobiological approach is that it lets you exclude all kinds of pests and nuisances from your space farm. The disadvantage is that your crops then grow in conditions that are unlike anything in nature or human experience, while still embedded in the relatively septic environment of a manned space station. In those conditions, what you thought was a harmless symbiotic bacterium might step up and become an opportunistic pathogen. Or you might discover that an organism you left out was a crucial symbiote necessary to the health of the parent plant.

And there’s evidence that plants and animals need a baseline level of pathogens to adequately develop. Sheltered plants might stand no chance in situations where a more grizzled, street-smart plant could easily fend off a challenge. Or they might simply not be able to produce offspring.

The risk of the ecological approach is also clear. You have no good sense of what organisms are living on your space farm, or how they might behave in an artificial environment. There is also the permanent risk of stowaways (or volunteers, as they were charmingly called by the Biosphere 2 crew). But at least you’re starting with a known balanced system and trying to keep it stable, rather than staking everything on a few chosen species adapting to an alien environment.

The difference in approach also affects your mindset. Are you an engineer or a farmer? If it’s the latter, you gain access to a very deep store of expertise and knowledge, but at the price of full control and understanding. If it’s the former, then you are hanging a giant KICK ME sign on your back to taunt Mother Nature, like every Silicon Valley startup that has so far ventured into vertical farming and found itself obliterated.

The interesting thing about Biosphere 2 is that it was built by a bunch of barely rehabilitated theater majors, but has so far blown all comparable attempts by European and American space agencies out of the water. Until someone can do better at closed system living, the burden of proof will be on NASA to show why hippie-in-a-van organic farming should not be the default approach for long-term living on the Moon or Mars.

Some basic numbers

With all this philosophical underbrush cleared, let’s build ourselves a Martian farm!

NASA says it takes 50 square meters to grow enough calories to feed one person. That area is more than twice what you need to generate oxygen and absorb carbon dioxide for one human, so in planning our farm we can treat food as the limiting factor, confident that air replenishment will take care of itself.

Biosphere 2 was able to adequately feed a crew of seven (and struggled to feed eight) on half an acre of farmland. That comes out to about 280 square meters per person. Much of the discrepancy with the NASA figure comes from crop failures (broad mites ate the potato crop, mildew got the soybeans) and insufficient lighting. In what follows, I’ll assume a figure of 100 square meters per astronaut, to allow for a safety factor and some variety in what is grown.

Sunlight on Mars is not bright enough for staple crops to thrive. Full Martian sunlight is about the equivalent of a cloudy day in the American heartland, enough for shade-loving plants, but inadequate for growing a proper crop of wheat, potatoes, or rice. The diagram below shows that even great Martian weather is marginal for potatoes, and if there’s any dust in the air, first the potatoes and then you are going to die.

Another variable of interest is atmospheric pressure. Experiments on Earth have shown that staple crops will grow normally in air pressures down to 10 kPa (about 1/10 of sea level pressure on Earth, the equivalent of 50,000 feet), as long as they are given enough carbon dioxide. But this is still about twenty times higher than the surface pressure on Mars. So a simple heated greenhouse is not going to cut it; you need some sort of pressurized habitat on Mars to grow stuff.

No Domes For You

Everyone wants a dome. I understand. I want a dome! The gorgeous SpaceX renderings of ‘Mars Base Alpha’, back before SpaceX became an AI hosting company, show a lovely transparent geodesic dome full of greenery. But this is an example of what I call the Utah Fallacy. Because Mars looks so much like the desert Southwest, there’s a tendency to underestimate the harshness of the environment there, and assume you can build desert Southwest-type things (like Biosphere 2!) right on the surface.

The trouble is that Mars is really more of a frozen, radioactive asteroid than a Mormon heartland. A geodesic dome full of Earth-standard air would act like a balloon squeezing with all its might against the Martian surface, trying to propel the dome high into the sky. The whole structure would have to be either bolted to bedrock or weighed down by many meters of rock and ice from above.

You can build such a dome on Mars (you can do anything!), but it’s hard to make a case for it. To build a sixty-meter dome, for example, you would need several hundred tons of carbon fiber struts and plastic panels2, along with another few hundred tons of bottled air to fill it3. This dome would initially leak at every joint, and the amount of work needed to seal it is completely unreasonable, even on Earth. De-leaking Biosphere 2 took months of sedulous effort (a lot of it was done by burning incense and following the smoke), and that was in a design where the pressure differential between inside and outside was tiny. Nor did anyone have to do the work while wearing a space suit.

A basic dome also has to be heated. If you’re one of those people who owns a vacation home that is mostly windows, you will be sympathetic to the heating problems on Mars. In full noon sunlight, a transparent 60-meter dome will hold an internal temperature of 23˚C, but at night or in dust storms it needs about a megawatt of power to keep the contents from freezing.

The worst thing is that you can’t even rely on the transparent dome for light. The solar flux on Mars is about 43% of what it is on Earth, and the dome structure will block another half of that. So you end up with less than 1/4 of Earth sunlight before factoring in dust storms, which can last for months and absorb 90% of incoming sunlight. This means that after all the work you did building a transparent dome, you still need basically the same number of grow lights as you would in an underground facility.

Those grow lights will also need power (although they help heat the dome!) and you can see how this all starts to get a little bit hard to assemble and maintain with robot labor 200 million miles from Earth.

The dome is also not a safe place to live. Besides the risk of sudden depressurization, there is the chronic radiation hazard from galactic cosmic rays. As I’ve written about before, the high-atomic weight component of this radiation poses a health hazard that is not captured by standard radiation models, so a crew living on Mars will have to spend most of its time hiding under several meters of rock to avoid getting all the cancer.

So what you’ve ended up building is an ornamental dome. At that point the sensible thing is to skip it entirely and move all of your farming into the underground tunnels you’re going to have to dig for your crew anyway.

To save on marsmoving equipment, it might make sense to find a lava tube or underground cavern that can be sealed off and converted to a pressurized space without all the bother of having to burrow into the soil. But the point is, for the first few decades at least, ain’t nobody going to be living in a dome.

Soil or Hydroponics?

So now you’re underground, looking out at the sealed tunnels you will spend the rest of your life farming. The next choice you have to make is whether to grow your crops in soil, or try your hand at hydroponics.

The argument for soil is easy. We have ten millennia of experience growing food crops in dirt, we know how it behaves, and what can go wrong. An important property of soil is its high inertia. It takes time and effort to heat it up, dry it out, get it wet, enrich or deplete it in nutrients, or make it do anything else. You can render your soil unsuitable for plants, but it’s hard to do quickly.

Read more

Analysis after the mission concluded that dehydration and lack of sleep had exacerbated an electrolyte deficiency that had probably set in during the final days of training before launch. NASA estimated that the lunar astronauts had lost about 20% of their body potassium (later work also implicated magnesium). Irwin experienced the more severe heart symptoms of the two, and it’s possible that his lunar exertion did him lasting harm; he died of a heart attack in 1991.

On the next moon mission, NASA overcompensated, fortifying the astronauts’ food with lavish amounts of potassium and pushing them to imbibe an orange drink laced with the element. An accidental open mic moment captured astronauts John Young and Charlie Duke discussing the effects of this diet while relaxing after a moon walk.

Young: I have the farts, again. I got them again, Charlie. I don’t know what the hell gives them to me. Certainly not...I think it’s acid stomach. I really do.

Duke: It probably is.

Young: (Laughing) I mean, I haven’t eaten this much citrus fruit in 20 years! And I’ll tell you one thing, in another 12 fucking days, I ain’t never eating any more. And if they offer to sup(plement) me potassium with my breakfast, I’m going to throw up! (Pause) I like an occasional orange. Really do. (Laughs) But I’ll be durned if I’m going to be buried in oranges.

NASA dialed down the potassium on the the final mission. But the heartbeat episode was just one of many dietary lessons the agency learned the hard way on Apollo. The moon landings were the first time NASA had to integrate meal planning into a large, complex mission, and there was a lot to get wrong. Perhaps the most important lesson was that Apollo food rations, a system of dehydrated pouches and rubbery cubes, were borderline inedible. A man who described himself as a “human garbage can” volunteered in 1969 to eat the Apollo diet as a test and declared that after four days, all the joy had gone out of eating.

He had regrets at the first meal, breakfast. The sausage patties seemed like flavored ground rubber and left a long lasting and sickening aftertaste. After one day, the tester’s appetite was much less than usual and by the third day mealtimes were unwelcome and eating was unenjoyable work. Space food similar to normal every day food was good but food designed for space was very bad. Chocolate and peanuts, which could have been left alone, were ground up and converted into bite-size cubes that stuck to the teeth.

To make matters worse, the water astronauts used to reconstitute their food contained large amounts of dissolved hydrogen gas, leading to bloating and more space farts. The Apollo crews also struggled with the mechanics and hygiene of zero gravity dining. In his memoir, Irwin describes a typical Apollo mealtime in some detail:

We had Velcro all over the spacecraft and on the meal packages, so to keep track of things we’d stick our dinner on the wall, course by course. If you nudged the meat course accidentally, it would take off, and you would have to float after it or get the help of a buddy downfield.

The soups were probably the most popular food that we had on board. We also had a variety of meats in aluminum foil—ham, frankfurters, turkey, and steak, in flat slices except for the frankfurters. The meats had different consistencies, but all had the same taste. When we cut open the end of the aluminum package and slid the meat out, the gravy or grease would slide out in blobs and float around the spacecraft [...] when we had finished eating, we had food all over us like a bunch of two year olds.

The food packages were very compressed when we started out, but after you finish eating the food you end up with packages that are twice as big, and you have the problem of stowing them away.

This involved cutting open another bag, breaking out our supply of yellow bacteria pills, and putting a few into each bag along with remnants of food. Then we would wind up each bag like a spent tube of toothpaste and shove them into the large plastic bag and pop that into a container. You had to squeeze everything real tight, or the bags would fill up with oxygen and billow out. We didn’t have room for that, and we didn’t have room to grow a crop of bacteria either. Sometimes, with all the bags around in different stages of the cycle, it looked as if the Monday wash was hanging out.

Housekeeping! Nothing but housekeeping!

Just like the space toilet, experience with in-flight dining on the moon missions showed that there were limits to human fortitude, even among the astronaut corps. So for the Skylab missions, which would last up to 56 days, NASA made a serious effort to make eating easier and more pleasant. The big lesson was to replace the tubes of featureless goo and freeze-dried powders with pre-cooked wet pack meals that bore at least some resemblance to normal food on Earth.

By far the biggest culinary hit on the Apollo missions had been a surprise Christmas dinner sent up on Apollo 8, which included a pouch of turkey and gravy. Unlike every other space meal served up to that time, it contained recognizable chunks of meat and was a sensation with the astronauts. NASA had developed and tested the technology in secret, as a treat for the crew, and it marked a path forward for space dining.

Skylab would prove much better than Apollo on the food front. It was also the only space mission where a crew ate its full allotment of calories. This was due entirely to cajoling from the medical team, who insisted on a strict accounting of food in and food out, with clean pouches at every meal.

Everyone agrees that the food on the International Space Station is tasty. Astronauts get to select their preferred menu on the ground, and like in a middle school lunchroom, there’s a lively orbital barter economy between astronauts of the various space agencies (Russian, European, Japanese, and American) who trade delicacies with one another. (The nuclear-strength Russian mustard is a particular favorite.)

The space station galley has approximately the same cooking facilities as an office breakroom. There is a microwave oven, hot and cold water on tap, and a very small refrigerator. (The coffee maker was removed some time ago after a successful trial run). Officially there is no alcohol allowed on board, but every once in a while NASA spends a few months trying to run down a spike in atmospheric ethanol that is inevitably traced to obscure origins somewhere in the Russian half of the station.

Astronauts on the ISS supplement their diet not only by trading with rival space agencies, but through periodic deliveries of fresh fruits and vegetables from Earth, as well as the occasional bit of salad or tomato from onboard experiments. Coolers needed to fly experiments down to Earth in a controlled temperature arrive on the space station packed with ice cream. And yet crews on the space station still consistently undereat, consuming only about 80% of their space calories on long-duration missions.

There’s no dearth of theories about why this happens. The ISS smells vaguely like a bus station toilet, the ambient noise level is high, there’s a lot of carbon dioxide in the air, and things in general taste different in space. Researchers have known for a while that flavors are altered even aboard commercial airplane flights. Sweet foods in an airline meal will taste less sweet, while umami flavors are enhanced, so passengers order more tomato juice and bloody marys in flight than they would ever consume in the airport lounge. Whatever the cause of the flavor change phenomenon, it seems to extend to space in an exaggerated form.

An ISS-like level of undereating poses a risk on Mars-length missions, where bone and muscle loss are already two of the most significant hazards of long duration exposure to free fall. The crew really can’t afford to lose any weight, and the hours of daily exercise they have to perform in space make the problem worse, by increasing their calorie requirement without adequately boosting their appetite.

So the challenge for Mars-length missions is to develop a food system that is safe, nutritious, and palatable enough to keep the crew well-fed for years, right down to the last meal they eat before returning to Earth.

Breaking the 5 year barrier

To understand why the Mars food problem is hard, consider the fate of a Mars-bound pizza. It begins life in an industrial oven, baked and then portioned into squares that get sealed into an individual foil bag. The bag is briefly heated (or perhaps zapped with radiation) to sterilize the contents, and from there the pizza is labeled, boxed, stowed, and begins its long journey through a logistics support chain that ends in warehouse near a launch facility.

One day, the pizza is placed on top of a rocket and launched into space, a shaky process that means nothing to a pizza but might disrupt a more fragile piece of food. If the slice is part of the meal allotment for a Mars-bound crew, it will spend the rest of its existence at room temperature (say 22˚C) and may have to wait two years or more before it’s eaten. If it’s being sent as part of a cargo shipment meant to be pre-positioned on Mars, it can make the trip over at a colder temperature (with no crew on board, nothing prevents cargo from being sent frozen) but will then have to spend three or more years in ambient conditions there, waiting for someone to come eat it.

Once you account for logistics, delays, and the short duration of launch windows to Mars, you can see how five years or more might elapse from the time a piece of space food is cooked and packaged to when it’s served for dinner.

There are three challenges to achieving such a long shelf life.

The first is avoiding spoilage, and for that we have no shortage of technologies. Foods can be pasteurized, microwaved, freeze-dried, frozen, irradiated, packed with preservatives, pickled, or canned. All of these techniques work well. Soldiers in Vietnam ate C-rations from Korea, Korean war soldiers ate surplus rations from WWII, and today YouTube is full of stunt diners sampling ancient military rations (all the way back to Napoleon!) with no immediate ill-effects, at least on screen.

Nutrition is a bigger hurdle. In addition to base calories and good quality protein, human beings need to consume thirteen dietary vitamins and about a dozen minerals. Some of these nutrients are indestructible, while others, particularly vitamins A, C, E and thiamine (B1), degrade quickly in stored foods.

Vitamins can be supplemented, but the supplements degrade as well. And there are non-essential nutrients, like phytochemical anti-oxidants, that would be good to have in the diet but are hard to store.

The most difficult problem in preserving food is palatability. Over time, all shelf-stable food wants to assume the color and texture of brown applesauce. This process can be slowed down by keeping food cold, but freezing has its own effects on texture, and it is impractical for a Mars mission1. The texture of shelf-stable foods degrades with age, and chemical reactions alter the taste and consistency of stored foods for the worse. This is not due to spoilage, but to interactions between the constituents of the food itself. The chief culprits in the process are the Maillard reaction and lipid oxidation.

Every home chef loves the Maillard reaction. It forms the golden brown crust on bread and gives seared meat its distinctive brown crust and depth of flavor. But in the food storage context, the Maillard reaction is a villain. While we normally associate it with high heat, it can also go forward in room temperature or even refrigerated foods, introducing off flavors and a sense of staleness. Milk products (including powder) are particularly susceptible, especially if they have been fortified with iron. Dried fruit is another frequent casualty. A notable problem with Maillard reactions is that they corrupt proteins, making them less digestible as a whole, while degrading the individual amino acids that constitute them.

Lipid oxidation is just a fancy way of saying that fats get rancid when you expose them to air. The amount of air this takes is tiny, making it a challenge to properly purge foods and exclude all air from packaging.

Both types of reactions go faster at higher temperatures, so the colder we can keep packaged foods, the longer their potential shelf life. But as I’ve written about in the past, bulk refrigeration is impractical on a spacecraft, which is best thought of as a large thermos.

So the global problem remains unsolved. Short of feeding a crew canned beans and ship’s biscuit, there is no known food system that can keep food unspoiled, nutritionally complete, and palatable for the five years necessary to sustain a Mars-bound crew.

What makes the problem especially thorny is the requirement for variety. If we could create a single balanced food (call it Soylent Mars) to meet the necessary nutritional and taste requirements, and feed a crew off of it, it would be one thing. But outside the world of software development, dietary variety seems to be a basic human need, with menu fatigue and concomitant undereating setting in quickly when an adequate selection of foods isn’t available.

Growing your own

Of course on Earth, we have an easy and cheap solution to the food problem. Plants make tasty food on demand, they recycle a number of astronaut waste products, and they can really liven up a spacecraft. So why not take some with us?

Read more

In November 2008, the Water Processing Assembly arrived on the ISS to realize the great dream of space exploration: boiling astronaut pee. The 800 kilogram Urine Processing Assembly would help take the station from a 45% to 80% water reuse rate. For the first time in the history of space flight, astronauts would be substantially recycling their water in an orbiting habitat.

In June 2010, thirteen months after the Water Processing Assembly went online, excessive levels of total organic carbon began to show up in the astronauts’ drinking water. Total organic carbon is a non-specific measurement that warns the crew about a contaminant being present, but gives them no clue to its identity.

When the space station was being designed, NASA had set the safety limit for total organic carbon at 3 parts per million, based on a worst-case scenario where formaldehyde got into the drinking water. By summer, the weekly trend in organic carbon was rising steadily and on track to exceed this threshold in December. At that point, NASA would either have to send up fresh drinking water or bring the crew back home.

There is no provision, then or now, for doing analytical chemistry on the space station. If you have a mystery substance, you need to put it into a returning Dragon or Soyuz capsule and wait for a lab on Earth to identify it. Astronauts and cosmonauts collect regular environmental samples, but whatever is in those samples only gets analyzed when that archive is brought down to Earth. So it wasn’t until September that a Soyuz capsule finally landed with the summer’s trove of water samples, which were quickly sent to the Food and Water Analysis lab in Houston.

There chemists confirmed the total organic carbon reading, but to everyone’s surprise couldn’t identify a specific contaminant in the samples. Whatever was in the water was not on the watchlist of several hundred chemicals that ISS engineers had anticipated might find their way into the station’s water system. In fact, the mystery substance wasn’t even in the lab’s vast reference library of mass spectra.

It took colleagues at Boeing, working from a newer reference library, to identify the mystery contaminant as dimethylsilanediol, or DMSD.

DMSD belongs to a family of compounds called siloxanes, molecules that contain a silicon-carbon-oxygen bond and occupy a kind of middle ground between organic chemistry and beach sand. Siloxanes (also called silicones) are common ingredients in cosmetics1, contact lenses, fake boobs, caulks, packaging, and all kinds of personal hygiene products, where they’re used to make things feel smooth and slippery. It’s siloxanes that give deodorant and hair conditioner their slick texture, and the same property makes them a popular industrial lubricant.

Manufacturers like siloxanes because they are cheap, stable, nontoxic, and unreactive, at least until they come into contact with something expensive aboard the space station. In ISS life support stories, siloxanes play the role of the meek character in an Agatha Christie novel who has been in the mansion the whole time, but who no one ever suspected had enough moxie to be the murderer. We will meet them again.

In this early brush with siloxanes, NASA greatly underestimated the compounds’ appetite for inflicting havoc. To confirm that DMSD was the culprit in the months-long excursion in organic carbon, chemists in Houston synthesized a pure reference solution of the stuff to calibrate against. They were happy to find that their state-of-the-art gas chromatograph/mass spectrometer was sensitive to DMSD, showing strong and clear peaks in every ISS water sample they looked at.

Unfortunately, the instrument also showed strong and clear DMSD peaks in everything else, including unrelated environmental samples from Earth and blank sample runs of distilled, deionized water. The chemists destroyed three expensive gas chromatographs before realizing that the tubing in their instrument was also made of siloxane. Once injected, the DMSD would happily dissolve into the walls of the chromatography tube and stay there, contaminating every future measurement the instrument made.

After devising an alternate analytical method that didn’t obliterate their lab equipment, the chastened chemists set to work figuring out how much of this DMSD stuff an astronaut could drink in a day without dying. They were still working on an answer when, to everyone’s surprise, total carbon readings on the space station dropped back to normal levels and stayed there, as if nothing had happened.

Since that episode, there have been at least five more spikes in total organic carbon in the station’s drinking water, all traced to DMSD. By NASA logic, this has turned siloxane contamination from a critical anomaly to a familiar behavior that can be modeled and planned against. The agency even boasts about its fight against siloxanes as an achievement of the space station, which is a little like bragging that your clifftop mansion helped further humanity’s understanding of erosion by falling into the sea.

Where do space siloxanes come from? Sleuthing has shown the main sources of siloxane vapor on the space station are antiperspirants, wet wipes, lotion, and leave-in hair conditioner. About a gram and a half of the stuff evaporates every day into the cabin atmosphere. There, helped by ionizing radiation from space, it decomposes to form the diol (DMSD), which is highly soluble in water. This compound collects in the water condenser, passes through the treatment chain mostly intact, and from there enters the clean water supply.

The dramatic spikes in total organic carbon observed aboard the ISS turned out to be a buffering artifact2. A more chemically active substance would bind to the ion-exchange medium in the water filtration beds and stay there. But DMSD binds weakly and can be kicked out by basically anything else. When a filtration bed is first installed, DMSD will begin to accumulate on the fresh resin, with no sign of it in the output water. But after some months, when the filter medium has saturated with DMSD, other substances will start to displace it, creating the signature rapid rise in organic carbon. Once all the DMSD that collected in the filtration bed has eluted out into the water supply organic carbon readings again drop to near zero. If a filtration bed is replaced, the process repeats.

Along the way, DMSD costs the space station a fortune. Each year a set of replacement multifiltration beds (which weigh 50 kilograms and have a three year design life) must be flown up from Earth.

NASA could try just ignoring the stuff. But there’s always a danger that DMSD-induced spikes in total organic carbon could be masking a rise in a different, more serious contaminant. And having this stuff in the output water does cause other problems. Siloxane has graduated to a known nuisance whose main effect is to shorten the life of the multifiltration beds in the water system and require the cabin heat exchanger (a 70 kilogram piece of metal) to be flown down annually to have its hydrophilic coating reapplied.

What makes dealing with siloxanes difficult is that they’re so inert. It’s easy to get reactive contaminants out of the life support loop, but siloxanes pass lightly through most of the various filters and ion exchangers. The only thing they seem to like to react with is catalyst beds and a costly and delicate hydrophilic coating on that heat exchanger. And when they do finally react, they do so by depositing a layer of glass, very effectively killing any reactive surface. It was a combination of DMSD and dimethylsulfone (from astronaut urine) that did in the space station’s experimental Sabatier reactor after only 1,800 liters of throughput.

After a fruitless search for some substance that might be able to sequester DMSD in water, NASA decided to attack the siloxane problem in the air phase, capturing the various siloxane vapors before they could hydrolyze into the diol and enter the water supply. In 2015, they replaced some of the rectangular HEPA air filters in the station with special siloxane-scrubbing filters packed with activated charcoal.

While these new filters reduced the concentration of siloxane vapor in the air, they also led to a mold outbreak. After running the new filter system for two and a half years, NASA has had to retreat to a hybrid solution—the filters are now half charcoal, half HEPA. This keeps mold counts down while capturing at least some atmospheric siloxane. At present the agency is testing a new filtration system to put in front of the heat exchangers, to try to protect them, and continuing to try to cut down on siloxanes at the source level. There are probably people at NASA now whose entire career has been built on siloxane control. But the status quo remains unsatisfying.

Siloxanes and Mars: a What-If

It’s interesting to imagine how the siloxane story would have played out if it was first encountered on a mission to Mars.

On the ISS, the initial rise in total organic carbon in 2010 came about 13 months after the station started recycling water. Assuming that every Mars-bound crew would be spending a few months on a shakedown cruise near Earth, that ten month mark would just about coincide with the ship’s arrival at Mars. There, the crew would put their spacecraft in a dormant state and move as a group to the surface habitat, with its own water recycling system, resetting the clock for the siloxane problem. A few months into their 17 month surface stay, organic carbon in the water would start to rise again, as DMSD started to elute from the ion exchange beds. Within half a year, the crew would face the difficult choice of whether to abort to orbit, swap out equipment, or accept the elevated levels of a mystery contaminant in their drinking water.

Regardless of their choice, once they got back to the orbiting spacecraft, they would see the rise in organic total carbon readings there resume, from the initial DMSD spike that was interrupted by their descent to the Martian surface. From the perspective of the crew, the same ominous problem would seem to be following them from orbit to the Martian surface and back.

If the astronauts had analytical equipment on board, like a gas chromatograph/mass spectrometer, the situation might grow even muddier. They would have trouble identifying the siloxane peak in their water samples, since it was not even listed in NASA’s reference database on Earth. And by the end of the hunt, enough DMSD would have dissolved into the tubing of their gas chromatograph to introduce a spurious signal to all future sample runs, causing endless potential confusion about its source.

A common pattern in aviation accidents is that efforts to fix a non-fatal problem rapidly snowball into a life-threatening situation, putting the crew in an unfamiliar part of the flight regime and eroding situational understanding to the point where they start to make serious mistakes. It’s easy to imagine how the siloxane issue, though ultimately harmless, might have prompted some bad choices on such a high-stakes mission with low margins for error.

Obviously any future Mars mission will have learned from the ISS experience with siloxanes. But the first encounter with the siloxane problem on the ISS is a good template for the kinds of problems we can expect to encounter on early forays away from Earth.

Actionable Business Insights From the Siloxane Affair

I like the siloxane problem because it such a mundane, annoying, and concrete demonstration of what makes life support hard. In particular, it illustrates how dependent space missions remain on large laboratories on the ground.

But there are broader LinkedIn-style lessons we can learn from siloxane!

Unknown unknowns are an underrated risk factor. There’s a phenomenon with Mars missions where the whole endeavor is so complex and difficult that it leaves little room for worrying about the unexpected. And of course it’s already hard to allocate mental space to things that you can’t describe or identify. But unknown problems are the ones with the most potential to seriously derail a space mission, since by definition they can’t be designed around or mitigated in advance.

Many unknown risks are unglamorous. The idea of facing the unknown has a nobility that’s hard to reconcile with the reality of deodorant sludge fouling the water supply. Siloxane vapor has been a known constituent of cabin air since at least the Skylab days. But the interaction with the water loop could only be discovered when someone made a serious effort at reprocessing water, and at that point a known bit player in the cabin atmosphere became a major programmatic headache. At no point in this mess were we making foundational discoveries about space travel, but it was a genuinely novel and hard problem.

Another good example of the great unknown being mundane is the discovery of space sickness during the Apollo missions. Today we know that about half of people who launch into space on large spacecraft3 will experience nausea and other symptoms of motion sickness during their first days in space, and that it is notoriously hard to predict who will suffer most. Utah Senator Jake Garn, a seasoned naval aviator with over 10,000 hours of flight time, was so afflicted on his sole Space Shuttle flight that an impressed astronaut corps adopted his name as the unit of spacesickness, reasoning that 1 Garn probably represented the physiological upper limit for human suffering. Today NASA manages spacesickness by giving it an acronym (Space Adaptation Syndrome) and letting astronauts barf it out during the first few days of their mission. But the unexpected incapacitation of veteran astronauts came close to seriously derailing several Apollo flights.

There was nothing particularly space-like to the nausea problem, just like there is nothing cosmic about siloxanes gumming up the water loop. But it took conditions that aren’t easy to replicate on Earth to stumble across the issue.

It is very hard to limit what gets aboard a large spacecraft. Once the siloxane problem was identified, an obvious strategy was to make a list of potential sources and exclude as many of them as possible from shipments to the space station. But NASA soon realized that siloxanes are ubiquitous and have so many applications that flagging them all is impractical.

There is a good cautionary tale here from the Space Shuttle era. That vehicle had heat resistant tiles that had to be attached to the aluminum belly of the orbiter. A special cloth had been certified for wiping the aluminum clean before applying the primer that securely bonded the tiles to the metal. After years of uneventful use, tile engineers discovered that new replacement tiles were no longer curing properly.

A careful investigation revealed that the supplier of that special cloth had changed the lubricant used in the machine that sews its hem. Minute amounts of the lubricant were being deposited on the stitching, and enough of that residue was getting on the aluminum skin to prevent the tile adhesive from curing properly.

And that lubricant’s name? Siloxane! I’m telling you this stuff is devious and absolutely hates spaceships.

Life support problems always find a way to compound. I’ve griped before that life support defeats engineering strategies for managing complexity by turning everything into a self-interacting tangle. Siloxanes are a nice specific example.

After discovering the siloxane problem, NASA tried hard to find a technique to selectively remove DMSD from cabin water. In the end they decided it was easier to attack the problem at its source, by filtering siloxane vapor from cabin air before it had a chance to decompose into the diol. The space station has large canisters that normally hold HEPA filters, and these could be re-purposed to try to trap siloxane. NASA tested a variety of filter materials to this end. Unfortunately, their top choice was so dense that the required blower fans would have exceeded the ISS limits for cabin noise. (The ISS is a cacaphonous place and those limits are a real health concern.) So the agency settled for a quieter and less effective filter made of activated charcoal.

But even this filter had an unintended impact. Astronauts soon started complaining of respiratory symptoms consistent with a high mold count. Those HEPA filters that had been replaced with siloxane-catching charcoal turned out to be not so replaceable. The ultimate solution was to split the difference and craft a filter cartridge that was half charcoal, half HEPA, which is the solution that the ISS still uses today.

What I want to highlight here is the complex environmental interaction between (1) space radiation, (2) trace contaminants, (3) surface coatings, (4) mold growth, (5) dimensional constraints, and (6) acoustic limits, with no easy place to break the chain. Radiation catalyzes a decomposition reaction in a trace contaminant in cabin air, polluting the water supply and damaging the surface of a heat exchanger. Attempts to filter out the contaminant run into volume constraints in the air system, and the easy solution (swapping out existing filters for ones better targeted at siloxane) turns out to have unacceptable repercussions for health. So the station limps along with a half measure that ameliorates the problem without solving it.

And this kind of thing is just a day in the life of a life support engineer.

Problems in space are hard to simulate. If you try to run a closed-loop water experiment on Earth, you will not find the water system getting gunked up with siloxane. Indoor air lacks the OH radicals that catalyze the hydrolysis reaction that turns siloxane vapor into the water-soluble diol. To find those radicals on Earth you need to go outdoors, where water vapor molecules get split by ultraviolet light.

But on the space station, there is enough ionizing radiation zipping around to split appreciable quantities of water vapor and drive the hydrolysis reaction in the absence of UV light. And so you get a problem in space that would never be caught in ground testing.

Further Reading

1. An early paper on the detective work required to figure out the siloxane-induced peak in organic carbon. The Story Behind the Numbers: Lessons Learned from the Integration of Monitoring Resources in Addressing an ISS Water Quality Anomaly (2011) DOI 10.2514/6.2011-5153

2. Part two of the tale: 2014 ISS Potable Water Characterization and Continuation of the Dimethylsilanediol Chronicle (2014)

3. A writeup on designing and installing the hybrid HEPA/charcoal filters that attempt to address the siloxane problem in the air: Design and Implementation of Combination Charcoal and HEPA Filters for the International Space Station Cabin Air Ventilation System (2019)

What fascinates me about sample return is that it’s a human mission in miniature. The goal is to land a large spacecraft on Mars and return its contents safely to Earth. The criterion for success is binary—either the sample gets back safely to Earth, or it doesn’t. The mission has a mix of American and European spacecraft launching over several synodic periods, interacting with equipment that has been pre-positioned on Mars years in advance. And the program as a whole consumes a substantial fraction of each agency’s exploration budget, spanning multiple election cycles.

This makes sample return a different animal from normal planetary science missions, which are set up so that there is a partial science return even if important components break, and whose costs tend to be front-loaded ahead of a single launch. For that reason, the demise of sample return may give us some clues about the political and organizational headwinds a human mission would face.

In my earlier post ,I described the various moving parts of the Mars Sample Return mission that died in 2026. Today I want to try swinging around the finger of blame, to see who it points at most comfortably. The various suspects have been assembled here in the drawing room. Let’s meet them one by one.

Donald Trump

Trump has a curious relationship with the nation’s space agency. He’s clearly interested in its achievements from a ‘national dominance’ point of view, taking obvious pleasure in NASA’s recently-completed Artemis II flight around the moon. Jared Isaacman, the agency’s new Administrator, has been sedulous about beginning every interview and public statement with praise of Trump. He’s also made getting a Moon landing in before the President’s term ends in 2029 the agency’s top priority. The flattery seems to be working. Trump keeps a model of NASA’s Mars-bound nuclear rocket on display in the Oval Office, brags about NASA, and seems happy with Isaacman.

But Trump’s proposed budgets for 2025 and 2026 both slashed funding for the science half of NASA, part of his administration’s broader jihad against Federally-funded science everywhere. Trump also set in motion personnel cuts that have cost NASA about 20% of its workforce.

Congress has pushed back against Trump’s budget cuts, keeping NASA spending fairly flat. But the political environment for science at the agency has been scarier. As the sample return mission became the most prominent and expensive line item in NASA’s science budget, there was internal concern that continuing to push for it would only start to cannibalize other programs. That zero-sum mentality removed a powerful internal constituency for the mission at NASA, and in the broader scientific community.

Bill Nelson

One problem with blaming Trump is timing. The sample return mission was already collapsing under Biden. So a more natural villain may be the man who oversaw the agency under the previous regime, NASA Administrator Bill Nelson.

Biden had the chance to nominate a dynamic new head of NASA. Instead he chose a fellow Senate dinosaur, 78 year old Florida Senator Bill Nelson. Nelson was a longtime space enjoyer best known for getting himself flown as congressional cargo on a 1982 Space Shuttle mission, back when NASA still indulged such stunts. At thta time, Nelson gained a reputation among the astronaut corps as a man enamored with his own reflection; his fellow crew members were warned never to get between the Congressman and a TV camera.

As NASA Administrator, Nelson treated the agency like a Cadillac Coupe de Ville on cruise control, a place where he could put his feet up and enjoy the sunset years of a long career in government. His grasp of science was minimal (at one press conference, he mortified his staff by talking about ‘the dark side of the moon’), but his love for the space program was pure. He was a man who loved a giant rocket and didn’t care who knew it.

Nelson mostly ignored the agency’s flagship science missions. His heart was with the Artemis moon program, and he spent his tenure making sure that Congressional money would continue to accumulate on NASA’s lunar bonfire, if necessary by cutting cheap and scientifically priceless missions like the Chandra X-Ray Observatory.

Nelson rarely mentioned Mars Sample Return until the program’s exploding budget started making headlines, at which point he called for commercial alternatives, then punted the problem to the next Administrator (and Administration). Nelson’s neglect didn’t kill the project, but it meant he missed the best chance to turn the program around.

Nelson’s biggest sin was insisting through his tenure that NASA was on track to put astronauts on Mars by the late 2030s. Anyone following the agency could see that even landing astronauts on the Moon in that time frame would be a challenge, and that Mars was unthinkable. But the timeline created an illusion of momentum and forward progress that bureaucrats at the agency found useful. Fantasy schedules are part of a culture of institutionalized mendacity at NASA that I’ve complained about before. But now that fantasy did real harm. As the earliest practical return date for Mars return moved past 2040, it raised a very reasonable question: why spend billions to return a half kilo of Martian soil in 2040 when arriving astronauts could just fill their pockets with the stuff in 2038?

Nelson’s refusal to give an honest answer— that we couldn’t possibly get astronauts to Mars before 2050 even with unlimited funding, and that at the current funding level we’re never going to Mars—painted him into a corner when it came to sample return.

NASA

Of course, NASA is a vast organization that is not just run by one man. The Administrator’s job is to fight for the agency on Capitol Hill and oversee its sprawling operations. Responsibility for individual programs rests further down, and this is where the leadership vacuum around sample return seemed most acute. No one seemed able to speak clearly for the program, or have full authority over its many constituent parts.

There were worrying signs that the institutional rot and budgetary incontinence that have plagued the human space flight side of NASA had now spread to the science side. For the second time in a decade, the agency had produced a Rube Goldberg mission architecture that seemed more about budget engineering than rocket design. And they seemed incapable of controlling either costs or schedule on a flagship mission.

The report from the second Independent Review Board was full of troubling evidence to this point. Key participants in the sample return program had never met in person, or else were working in the same physical location but reporting up different chains of command, sometimes to NASA centers on the other side of the country. Several of the people ultimately responsible for major components had no line authority over the engineers designing them. Even the Jet Propulsion Laboratory, normally a bastion of competence, had made uncharacteristically sloppy design errors with a rover’s landing legs that cost a fortune to rectify.

All this organizational Bedlam was reflected in the hardware. The part of the design that ‘breaks the chain’ between Mars and Earth by sterilizing the outside of the sample container couldn’t seem to stabilize. Initially, the Earth Return Orbiter was supposed to bake the sample container with heat. Then it was decided to sterilize the sample container with ultraviolet light. Finally the sterilization step got moved back to Mars, once again using an oven instead of lamps. Major changes like this had expensive, cascading effects across the program.

At times entire pieces of the architecture would wink in and out of existence. The fetch rover initially delegated to the Europeans turned into a pair of helicopters, then disappeared entirely. Core mission constraints (like the mass of the orbiting sample) kept changing. Conway’s Law says that every complex system reflects the organizational structure of the group that builds it, and NASA seemed determined to prove it in the most perverse way possible.

The organizational chaos was accompanied by a lot of lying. The exploding budget for the program was kept a secret until it could no longer be disguised. The official schedule was never updated, even after everyone involved knew it would be impossible to meet. The assumption was that Congress would do what it always does when the human spaceflight side of NASA steps into quicksand. There would be grumbling, hearings, and then a reluctant multi-billion dollar appropriation.

But there’s no Senator from Mars in Washington. Without the sacrosanct workforce that shields projects like SLS or Orion from outright cancellation, the program had no adequate political cover. And so Congress called NASA’s bluff.

Starship

Starship may seem like an odd choice for villain on a NASA science mission. But with initial flight tests of the rocket going well, and Elon Musk promising that enormous quantities of cargo would be landing on Mars by 2020, the tension with the sample return mission was real. Why spend billions trying to figure out how to land a minimum viable rocket on Mars in 2028, when Starship would soon be flying large payloads to that planet practically for free?

Given Musk’s apparent sincerity and limitless resources, it was taken as a given that SpaceX would eat the cost of learning how to land Starship on Mars. At that point NASA might be able to contract with the company to send its own heavy payloads there. Something similar had already happened with Artemis. After NASA blew its entire lunar budget on a rocket, leaving no money for a lander, SpaceX stepped in and offered to build the lander for a relative pittance. Now it looked like the trick might work again.

One wrinkle in using Starship for sample return was that the cached samples lay in Jezero Crater, a site of high astrobiological importance. Like all large spacecraft, Starship is impossible to sterilize and would land on Mars covered in filth. But these problems seemed like problems worth having. With Trump in office, it was possible Musk would just ignore the contamination issue outright. And even if responsible adults could keep Starship out of Jezero crater, there had to be some technical way to finesse the issue (helicopters! fetch rovers!) and get sample return to work.

The point was, a torrent of heavy cargo deliveries to Mars was coming. NASA could either get on the Starship train, or risk looking ridiculous when it spent billions of dollars to bring a volleyball full of rocks home more than decade after Starship had made two-way transit to Mars cheap and reliable.

Covid and Putin

Putin’s contribution to the sample return fiasco was derailing the timeline for the European Space Agency’s Rosalind Franklin rover. This homeless rover had spent years looking for a rocket to launch on before entering a partnership with the Russian space agency in 2013. The Russians had rockets but no money, while the Europeans had a rover and no rocket. The ESA and Rosskosmos were on track to launch their mission in 2020, but the pandemic pushed the launch date forward into 2022. In February of that year, Russia invaded Ukraine, making it politically impossible to continue the joint effort.

The loss of Rosalind Franklin was awkward, because that mission was supposed to give ESA the practical experience it needed to design a fetch rover for Mars sample return. Now the Americans had to build the rover, and by the law of flat budgets, that meant the sample return mission would be delayed by another few years.

As for covid, it got a lot of people who really needed to be working in physical proximity accustomed to working from home. The problem was not unique to NASA. As a commenter on the NASA Space Forum observed, no major aerospace project (civilian or military) has been able to hold to its schedule and budget since the pandemic. But the situation seemed to particularly affect the design teams at JPL.

Mars

No list of potential villains for sample return can be complete without Mars. The planet is cold, distant, and annoying to explore, with just enough atmosphere to make landing a hassle, but not enough to be helpful.

Two problems in particular bedeviled the sample return mission.

Landing heavy payloads. Every American rover and lander since the 1970’s had used an aeroshell design first developed for the Viking landers. NASA qualified the design at vast expense because it needed a way to land a pair of 800 kg spacecraft on Mars. Since then, the weight limit has been pushed out to two tons (with the Perseverance rover) and could possibly be stretched to three. But that’s about the outer limit for payloads, which means that every object landed on the Martian surface has had to be lightweight and built with the precision of a Swiss watch.

What we need is a way to land dumb, heavy payloads on Mars. Then vehicles and sensors wouldn’t cost so much to develop, we could tolerate a baseline rate of failure, and instead of artisanally hand-crafting each Mars probe, we could just crank them out serially. Sample return looks a lot different if there are multiple rovers going around collecting samples, caching hundreds of sample tubes, and making it okay to lose a few of those samples in the attempt to get them home. Larger payload capacity also clears the way for landing big rockets on Mars, including ones big enough to fly directly to Earth without performing an elaborate orbital pas-à-deux with an expensive Earth return vehicle.

The Viking-era constraint exists because it is extremely expensive to test and verify a system for entry and landing on Mars. Before this year, it looked like SpaceX was about to spend its own money figuring this problem out for everyone else. The company had the experience in heat shields and supersonic retropropulsion, were determined to start landing Starships on Mars in great numbers, and were led by a Mars obsessive with limitless financial resources.

But with Starship now out of the picture for Mars, someone has to step up and figure out how to land large (> 50 metric ton) payloads on the Martian surface. Otherwise surface operations there will never get cheaper, no matter how much launch costs go down on Earth.

Precision landing. This capability to some extent complements the one above. The Martian atmosphere holds a lot of dust, which heats the air around it and introduces uncertainty in where a payload might land on any a given day. This makes typical landing ellipses tens of kilometers long, and a couple of kilometers wide. There’s no way to land something in a precise location, or next to equipment that is already in place.

If we could instead land payloads with precision, the problem of being stuck in those Viking-size aeroshells would be less acute. Individual pieces would still be limited by the constraints of an aeroshell, but there wouldn’t be such a need for fetch rovers or landers sized for a worst-case aiming error. One could even try bold things like landing a return rocket in pieces and assembling it on Mars.

The Wolf Amendment

Our final villain, the Wolf Amendment, is a piece of legislation that forbids US space agencies from working with China. Chinese nationals aren’t even allowed to tour places like Kennedy Space Center.

It’s an interesting restriction. Years of sanctions and a proxy war in Ukraine have not kept NASA from sharing a space station with the Russians, or launching cosmonauts next to astronauts from Russian launch pads. The agency even used to boast about how cooperating on the International Space Station was helping ease geopolitical tensions.

But a lot of people want there to be a space race with China, so that Congress will let space nerds cosplay the 1960s. This attitude comes complete with ludicrous visions of a Red Moon, or China somehow seizing all the lunar water before we have a chance to get to it and… do what? Drink it? Make ice sculptures? The China threat is part of a strategy to manufacture political support for an otherwise pointless manned lunar program.

This is a shame, because there are such sound reasons to work with China on Mars! Both countries have an active exploration program and need the same things—high-bandwidth communications, a way to land heavy payloads, and precision landing capability. China is working on a rudimentary sample return mission that stands a good chance of success. The US has a set of scientifically valuable samples on Mars that it can’t get home. And if the countries worked together, they could create Apollo-Soyuz style good vibes. Instead of cosplaying the 1960s, we could be cosplaying the 1970s. That has to be some kind of progress!

The Verdict

It’s entirely NASA’s fault. Congress gave the space agency billions of dollars and years to bring a sample home from Mars, and NASA flubbed it. All they had to do was fly a single Martian pebble back to the vicinity of Earth, where scientists with cyclotrons and fancy microscopes could have swarmed all over it. If those investigations had found the smallest trace of life, rivers of money would have then flowed into Sample Return II.

The NASA of the 1970s could have pulled this off! But instead, the project disappeared into the bureaucratic haze that seems to have eaten the agency from within. For all the supporting villains I listed, it’s our own space program that no longer seems up to the task of actually exploring space. For the first time, a technical challenge proved insurmountable given ample time, budget, and manpower. And the effort failed largely for organizational reasons. That seems like an ominous sign for the future.

Subscribe now

Opinion around Starship has a tendency to polarize. On one end is an extreme enthusiasm and a conviction that Starship will revolutionize space flight by bringing launch costs down by orders of magnitude. The rocket’s fans point to SpaceX’s immense success in operating the Falcon 9, the world’s first reusable commercial rocket, along with the rapid progress Starship showed during early flight tests in 2023. Launches in that ‘V1’ configuration showed that Starship could reach orbit1 and successfully re-enter the atmosphere. Both stages of the spacecraft (the Super Heavy first stage and the Starship upper stage) demonstrated precision controlled landings, and in 2024 SpaceX was able to catch and re-fly a Super Heavy first stage. From the optimists’ perspective, the Starship concept has been proven; what remains is a series of iterative improvements to make it operational, followed by scaling.

The opposing tendency is a skepticism that can shade into derision. Starship dislikers point out that the vehicle has yet to reach orbit, and that Elon Musk’s promises about its payload capacity keep getting pushed to future versions of the rocket. No Starship upper stage has made it through re-entry in anything like a condition to fly again, even though rapid reusability is essential to the economics of the program. And the lunar variant of Starship, paid for with taxpayer money, is years overdue. Moreover, there has been a tendency for the whole rocket to get redesigned just as things start to stabilize. The criticism is that SpaceX is not doing iterative flight testing so much as YOLO flight testing, and that it can’t get its ambitious design to close.

I have a similar struggle evaluating Starship that I do with AI. The core technology is undeniably real and transformative, but it comes welded to a preposterous vision of the future. In the case of Starship, that means hundreds of launches a day, vast orbiting data centers, and kilometer-length mass drivers on the Moon built to sub-micron tolerance. And the whole thing depends on the emergence of an orbital economy that, for several decades now, has resisted boosters’ efforts to will it into being.

It doesn’t help matters that people talking about Starship are often talking about different things. Starship is simultaneously a very big rocket, the second stage of that rocket, an operational concept, an enabling technology for colonizing Mars, and a deus ex machina for solving all practical questions of human space flight.

But between the reality and the woo there is a gray area that is hard to evaluate. So I’m writing this post to try to clarify my thinking, and in particular to try to figure out at which station I want to get off the hype train. I always welcome comments on my posts, but I would particularly welcome them on this topic, since there are so many different ways to think about Starship.

Starship as a vision

Starship was born as a Mars-going rocket, and the argument for it then was simple. Launch costs were the biggest barrier to settling Mars. The only way to make rockets cheap was to mass-produce them and stop throwing them away after each launch. So SpaceX set out to build a very big rocket that was rapidly re-usable and could ferry heavy payloads to the Martian surface.

SpaceX was serious both about the ‘rapid’ and ‘big’ parts of this vision for Starship. Their targeted payload to low Earth orbit was 100 metric tons, or about the mass of a fully-loaded Space Shuttle orbiter with a satellite stuffed inside.2 This capacity would make Starship more capable than NASA’s budget-obliterating Space Launch System, and nearly as capable as the 1960’s vintage Saturn V. A Starship that refueled in orbit (another part of the vision) could then go on to land that 100 ton payload on the Moon or Mars. When you consider that the heaviest object ever landed on the Moon weighed 8 tons (Apollo 17), and the heaviest payload to reach Mars weighs one ton (the Perseverance rover), you see the transformative potential of the technology.

Operationally, the giant rocket was supposed to behave more like an airplane, able to launch again within minutes after landing, and to fly many times before requiring a major overhaul. This capability would be a step change from the Falcon 9, whose reusable first stage has to be overhauled for days or weeks between launches, not to mention a dinosaur like the Space Shuttle orbiter, which needed a full rebuild between launches.

As a final touch, Starship would use methane and liquid oxygen as rocket fuel, opening the possibility of refueling the rocket from (hands begin waving) autonomous solar-powered propellant factories on Mars.

So the original Starship vision consisted of four elements:

1. A huge, cheap rocket

2. That is rapidly reusable

3. And can refuel in orbit

4. Or on the surface of Mars.

Let me go through what’s left of this vision in descending order of plausibility.

Super Heavy

Super Heavy, the enormous first stage powered by 33 Raptor engines, has somehow become the least contentious part of the Starship design, despite looking like a mad rocketeer’s fever dream, a cluster of giant engines tucked under a skyscraper of rocket fuel. The only point of reference for Super Heavy before it was built was the Soviet N1 rocket, a mighty vehicle that blew up on all four of its launch attempts and discredited the idea of ‘just add engines’ for a generation. But SpaceX made it work, almost from the first launch.

Today Super Heavy is a piece of pure electric-guitars-and-screaming-eagles space awesomeness, from the gorgeous purple exhaust plume full of shock diamonds, to the grace with which it descends on a single swiveling column of flame until the rocket comes to rest on the chopsticks that catch it.

The fact that SpaceX seems to have had less trouble with this vast booster than the simpler upper stage was one of the big surprises of the flight test campaign. While Super Heavy had some early engine failures on the first two flights, these were ironed out quickly. The fifth and seventh test flights demonstrated a successful return of Super Heavy to the launch pad, and the booster from that second catch was then re-flown two flights later.

If SpaceX was using Super Heavy as the first stage in a more traditional rocket stack, there would be no reason to write this post. Both the booster and the Raptor engines that power it are a giant leap forward and the main reason it’s so hard to dismiss ambitious promises from SpaceX about the rest of the system.

The Second Stage

The second stage of Starship, confusingly also called Starship, is where SpaceX has struggled. The problem is mostly one of size. In order for Super Heavy to fly back to its launch site, Starship-the-rocket has to stage very early, which pushes a lot of mass and propellant into the second stage. That stage is about as tall as the Space Shuttle and its attached fuel tank put together, and when fully fueled, nearly as heavy as the entire Space Shuttle stack was when it sat on the launch pad.

Keeping this hunk of spacecraft intact and controllable during re-entry requires large control surfaces and a capable heat shield. But for these elements to be re-usable, they have to be sturdy, and ‘sturdy’ in an aerospace context usually means so heavy that it eats through all your available payload.

The early Starship flight tests drove home the difference between a rocket that can re-enter once, and one that can survive re-entry dozens of times. On several of those flights, viewers could watch hot gas slicing through the steering flaps in real time, even as the hardy rocket shrugged it off and came to a successful water landing. Some landings left visible damage on the engines, and all of them left the vehicle covered in spectacular burn scars. To what extent the damage is structural, and to what extent it is cosmetic, only SpaceX knows. But it shows the magnitude of the forces at play.

The nightmare scenario for SpaceX is one where making Starship robust enough to survive repeated atmospheric entries leaves it too heavy to carry useful payload. This seems to be what happened with V1. That first iteration of Starship was supposed to deliver 100 tons to low earth orbit, a claim Musk later walked back to 45 tons, then ultimately 15 tons, less than the payload of the Falcon 9. The V2 version of the rocket (a horrible version number to use in rocketry, just saying) managed to lift 35 tons to near orbit, but was also prone to explode, accelerating the shift to the new design that launched on Friday. Starship V3 has incredible new engines and looked peppy on its maiden flight, but it’s not clear how much mass will have to be added back to the rocket before the latest design either stabilizes or gets ditched for an even more ambitious V4.

Put simply, SpaceX is in a race to see whether it can improve performance in the Raptor engines faster than the upper stage of Starship gains weight. The real test for the program will be the first capture and re-flight of an upper stage, because the economics of launch cost are sensitive to just how many times that upper stage can fly. If it turns out a Starship upper stage is only good for a flight or two, the rocket will remain groundbreaking, but cannot meet its more transformative goals.

Operations

How often will Starship launch? Getting a credible answer is hard because the SpaceX IPO hinges on preposterous numbers that the company can’t disavow yet.

The lower limit to launch cadence is set by SpaceX’s contract with NASA for a lunar lander, which the agency calls the ‘Human Landing System’. HLS has to be refueled from an orbiting depot Starship to get to the Moon, and filling that depot will involve launching a large number of tanker Starships over the span of a few months. How many launches that will take is unknown; I think people would be surprised to see it happen with fewer than ten launches, or more than thirty. Assuming that a fuel depot can hang around in orbit for six months, that implies an operational tempo of about one launch per week, from at least two different launch sites (so that a pad explosion at one won’t ruin the whole campaign).

This would be an ambitious but achievable cadence for Starship, particularly if the tanker Starships didn’t have to be reusable, and could be made without a heat shield or control surfaces. Just how much time SpaceX has to reach this cadence is an interesting question, since right now NASA and SpaceX are locked in a game of chicken around who can commit to the least realistic timeline. But it probably has to get in gear by early 2027.

A more ambitious goal would be to match Falcon 9’s cadence of one launch every two days. At that high a launch rate, Starship would also need to match Falcon 9’s high level of reliability (about 1 failure in 200), and expand launch operations across several more sites. SpaceX would also need to find payloads to put on all those rockets, unless you accept their pre-IPO claim that the appetite for Starlink and AI launches will be insatiable for decades to come.

Finally, there is the launch cadence SpaceX actually targets in their S-1, a million metric tons a year to Earth orbit. That frankly preposterous figure implies 25-30 Starship launches a day, with the exact number contingent on how much payload the final version of the rocket can carry. This would be the flight rate of a small regional American airport like Chattanooga or Sioux Falls, except that instead of turbojets SpaceX would be launching skyscrapers full of liquid oxygen and methane from a constellation of launch pads, each one handling multiple launches, catches, and re-stackings in every 24 hour period.

Launching Starship on the hour would also mean permanent no-fly zones for aircraft and a likely environmental backlash against SpaceX, who would be putting significant amounts of water vapor in the stratosphere. Overnight the company would become one of the country’s biggest consumers of methane, electric power, and liquid oxygen. And since a failure rate of 1/200 at this cadence would have Starships falling out of the sky every week, the rocket would have to improve in reliability by at least two orders of magnitude.

In other words, SpaceX’s IPO promise implies a Starship program that looks like airline travel in the early thirties, dozens of flights a day with an accident rate of a few flights per hundred thousand, or about a thousand times more reliable than any modern rocket.

However much you may love SpaceX, there is no number of bong rips that makes this scenario feel real. It’s in the S-1 is to try to prop up the company’s astronomical valuation, but the sooner we can all move past it, the better.

Orbital Refueling

At this point we leave the realm of existing hardware and get speculative.

Back when going to Mars was the goal, orbital refueling was a key part of the Starship concept. The heavy Starship second stage arrives in orbit with its propellant tanks almost depleted, and to travel to a destination beyond Earth orbit, it needs to top up. This procedure—pumping cryogenic propellants between two rockets in orbit—has never been attempted.

Now that SpaceX is an AI company, refueling has become something of an awkward side quest. SpaceX can launch as many Starlink and AI satellites it wants on Starship without having to touch refueling with a barge pole. But it is still on the hook to develop the technology for its NASA contract, since the lunar lander version of Starship can’t get to the Moon without it. And turning refueling from an idea to an operational capability has the potential to get expensive, especially if the company is not yet able to re-use Super Heavy and second stages while launching all those tankers.

Multiple factors conspire to make the refueling problem hard:

• Historically, valves and seals are two components that love to act up in space. But for refueling to work, two rockets have to come together, form a leakproof seal, and move hundreds of tons of propellant through some elaborate plumbing.

• Propellants in zero g exist as a three-dimensional jumble of liquid mingled with gas. Moving them between rockets requires either developing special wicking techniques for use in zero g, or accelerating the docked rockets enough that the propellants inside settle to the bottom of the tank.

• Liquid methane and oxygen are both very cold, and the plumbing connecting the two rockets has to be chilled down by spraying it with propellant that flashes into gas. Doing this with minimal propellant loss needs careful geometry and is difficult to model with computers.

• When Starship and the fuel depot dock together, the propellant inside will constitute a significant portion of the two vehicles’ mass. As that propellant moves from Starship A to Starship B, it will shift the center of gravity of the combined assembly. This creates challenges if the rockets are being accelerated in tandem to settle propellant in the back of the tank for pumping. It also creates the potential for slosh as moving fluid bounces off the interior of the propellant tanks.

• Once propellant has been transferred to an orbital depot, it has to stay there for a long time (weeks or months) without boiling away. Keeping cryogenic gases liquid while a warm planet fills a third of the sky requires active cooling and careful pointing.3

Just like with the Starship heat shield, no one doubts that orbital refueling is possible. The question is one of efficiency: how many Starships have to launch from Earth to fully fuel a Starship that wants to leave orbit? At what point does the sheer number of launches required make the whole idea uneconomical?

NASA’s official schedule for the Artemis missions anticipates landing a crew using HLS in late 2028. Working back from that date, SpaceX would need to have orbital refueling operational by the end of next year, giving the company 18 months to complete a sequence of increasingly demanding test objectives:

1. Complete an orbital test flight of Starship.

2. Launch two Starships at once and demonstrate that they can rendezvous in orbit and dock together.

3. Transfer some quantity of propellant between two docked Starships in orbit as a proof of concept.

4. Transfer a full load (about 100 tons) of propellant from a freshly-launched Starship to an orbiting Starship configured as a fuel depot.

5. Keep the fuel depot in orbit for several weeks to test station-keeping and determine the boil-off rate of stored propellant.

6. Demonstrate a full orbital transfer of propellant (~1200 tons) from the orbiting depot to an empty HLS Starship.

Fortunately for SpaceX, NASA’s timeline for landing on the Moon is even less realistic than these development milestones. But it does set up an interesting dynamic of who will blink first, and who will bear official blame for the inevitable delay of the first Artemis landing into the 2030’s.

Mars

The final and most ambitious part of the Starship vision is landing on Mars. Back in the innocent days of 2025, Elon Musk promised that flights would start this year, with serious tonnage starting to arrive on the planet in 2029.

But it’s hard to read SpaceX’s S-1 filing and estimate the chances of a Starship landing on Mars as anything higher than zero. The company’s core business is no longer space flight, but data center rentals and B2B enterprise sales in low Earth orbit. Not a penny of the company’s claimed total addressable market of $26 trillion(!!) comes from sending space nerds to Mars.

This is bittersweet news. The bitter part is that before the xAI merger, SpaceX seemed legitimately committed to trying to land Starship on Mars. In an earlier post, I wrote about how every lander since Viking has been trapped in the same basic aeroshell design, making it hard to put more than two tons of equipment on the Martian surface at a time. Starship was our best chance at breaking that barrier, and it looked like SpaceX was even going to develop the capability on its own dime. NASA faced the real possibility of being able to land a hundred tons of stuff on Mars for about what it costs to send a single rover there now.

Unfortunately, every Starship that launches into space is also a big piece of machinery that stands out in the Texas sun. Bugs crawl on it, water collects in it, and bacteria, mold and fungi cover it from tip to stern. Like any large spacecraft, Starship is much too big to sterilize. And some of the creatures that might take up residence on it are unfortunately durable enough that they would survive a six-month trip through space, making every Starship that landed on Mars a severe contamination risk.

As a practical barrier, this meant nothing to SpaceX. Musk has been clear that he has no interest in the contamination problem, and it’s hard to imagine a Trump-led FAA withholding a launch license from him to protect the sanctity of the 1967 Outer Space Treaty. But for people who dislike contaminating Mars on the merits, sending Starships to land there was a Faustian choice. On the one hand, it would enable large-scale robotic exploration, but on the other hand, it might vitiate the whole purpose of sending all those robots in the first place.

But now—bittersweet! There is no more Mars plan. We’re back to the pre-Starship struggle of trying to figure out how to land heavy stuff on the planet, but at least nobody is trying to bury giant septic rocket hulls there to live in.

Space Shuttle II?

Six months ago, the SpaceX story was clear. The company’s long-range goal was to colonize Mars, and its lucrative launch business and Starlink internet service existed to subsidize the development of Starship, a rocket that would transform the economics of heavy launch and make the Mars plan real.

Today, the story is confusing. SpaceX is suddenly an AI company with a small sideline in rockets, claiming a $2 trillion valuation on mostly vibes. Starship is no longer a rocket for colonizing Mars, but has turned into what the Space Shuttle was trying to be early in its design—a cheap, reusable two-stage space truck.

I can’t imagine Starship failing in this diminished role, even if making the upper stages fully reusable turns out to be too expensive. SpaceX is bound to get some variant of ‘Space Shuttle’ Starship working, lowering its launch costs substantially, and fulfilling its dream of data centers in spaaaaace.

The interesting question is what happens with orbital refueling. Over the last half decade, NASA had engineered an enviable situation where Elon Musk and Jeff Bezos were both willing to go out of pocket to develop a transformative technology, just to stick it to the other guy. The nation’s space agency had finally found a way to harness the power of our space billionaires. But with SpaceX pivoting from Mars to AI, orbital refueling is no longer on the critical path for the company’s goals. And Blue Origin, the other player in the refueling game, might also decide to abandon those efforts in favor of launching technically simpler transfer stages.

If SpaceX decides to follow through on its HLS commitment, and finds a way to make refueling routine and economical, then that really will open up the Solar System to exploration. We will be able to cheaply send probes and rovers to the outer planets, darken the sky with space telescopes, and build significant infrastructure on the Moon and Mars. But if refueling turns out to be impractical, or too expensive for commercial applications, then Starship will end up a less transformative technology than some of us space nerds had hoped.

Either way, though, what an impressive piece of hardware! If SpaceX hadn’t built it, people would call it impossible, and that’s the highest form of praise for the engineers who have beavered away on this rocket for so many years now.

Subscribe now

Further reading/viewing

• Scott Manley always has nice Starship recaps, and his video going over the latest test flight is no exception.

• The SpaceX S-1 makes for some amazing reading, even if you normally don’t curl up at night with SEC filings.

• This 2024 paper is probably the most detailed attempt to model a Starship Mars mission from the outside. About feasibility of SpaceX’s human exploration Mars mission scenario with Starship (2024). DOI 10.1038/s41598-024-54012-0

This claim touches on the very interesting question of what kinds of life we can identify on Earth.

As early as 1932, the Soviet microbiologist A.S. Razumov noticed that there were far more bacteria visible under a microscope in water samples than could be persuaded to grow in his laboratory. In a 1985 paper, the researchers Staley and Konopka christened the discrepancy ‘The Great Plate Count Anomaly’. By their estimate, something like 99% to 99.9% of bacteria in freshwater samples were happy, metabolically active microbes who could not be bullied or bribed into growing on a Petri dish or in a vial of liquid medium.

Back in an era when culturing microbes was the first step in trying to understand them, this was a problem. The only thing researchers could do was cross their fingers and hope that the few species we managed to grow in the lab were representative of the great unwashed masses of microbes living in the world around us.

In 1977 Carl Woese, using novel molecular techniques, announced the discovery of a third domain of life he called Archaebacteria. Later renamed Archaea, these one-celled organisms looked like bacteria under a microscope, but were no more related to them on the molecular level than they were to a rose bush or a hamster. It took until the 1990’s for Woese’s three-domain model of life to be accepted, but the methods he pioneered immediately transformed the field.

Woese reached his landmark conclusion by examining ribosomal RNA. Ribosomes are an ancient piece of cellular machinery, common to every form of life on Earth, and the DNA that codes for them was short enough to be tractable even by 1970’s standards. The ribosomal DNA had a useful property: some sections of the sequence are highly conserved, because key parts of the ribosome can’t function without them, while others are less constrained and can mutate freely over time. The existence of the conserved sections made it easy to pick out ribosomal DNA from a sea of fragments, while statistical analysis of the mutable areas gave a measure of evolutionary distance.1

Woese had used ribosomal DNA to reveal relationships between microbes that could be grown in a lab. But a group led by Norman Pace quickly realized that the same technique would work on microbial populations in the wild, without having to culture anything. The line of research his group opened over the next ten years changed the face of microbiology, uncovering not just new species, but entire new divisions of life. Woese had grouped the known bacteria into about a dozen phyla, and now Pace’s group added twenty four major categories they called ‘candidate phyla’—new domains of life that we knew existed, but had no additional knowledge about. It was like opening a door in your studio apartment to discover that you’ve been living in the closet of a dark and unexplored palace.

At this point microbiologists faced a problem more familiar to mathematicians. Mathematics has the notion of a non-constructive proof, where you demonstrate the existence of a mathematical object without being able to produce an example of it. The ribosomal studies proved the existence of whole new categories of microbe, but didn’t offer clues about how to find such a creature in the wild. These new arrivals remained spectral entries in a growing database of ribosome sequences.

You can think of cataloging ribosomal DNA like being told someone’s name, date and place of birth. It’s a helpful way to build a family tree, but tells you nothing about a person’s appearance, hobbies, personality, or behavior. What was needed was a way to connect the new phylogenetic categories to actual organisms that could be collected and studied.

The next step was to try sequencing entire genomes directly from environmental samples. This technique, called shotgun sequencing, would have been ruinously expensive in the 1990’s. The Human Genome Project had needed thirteen years and nine-figure sums to assemble a rough draft of the 3 billion base pair human genome. But the cost of automated DNA sequencing was dropping at about the same pace as Moore’s Law, with the price per base pair cut in half every 18 months. In 2004, Jill Banfield’s team proved that environmental shotgun sequencing was viable by isolating multiple genomes from a sample taken in a California iron mine. That same year, Craig Venter collected samples from across the Sargasso Sea and assembled a list of 1800 candidate species. Their experimental techniques improved quickly; Banfield would go on to discover a bestiary called the Candidate Phyla Radiation, which for a while looked like it might constitute a fifth of all microbial species on Earth.

The floodgates really opened after 2005, when a new generation of commercial sequencing technology came online. The per-base cost of sequencing dropped by something like five orders of magnitude in just four years. You no longer needed a big lab and a million-dollar budget to try ‘shotgun sequencing’; all it took was a graduate student and a shovel. The field entered a sort of Wild West phase where you could put random pocket lint in a 454 pyrosequencer and reliably discover not just new species or genera, but entire phyla. Discoveries at this level were unheard of in multicellular biology—imagine if it had taken us until the year 2000 to discover insects or flowering plants. But in microbiology they were now expected.

A final breakthrough came with the adoption of multiple displacement amplification, a technique that could take DNA from a single cell and grow it into quantities big enough to be sequenced. Instead trying to tease out individual genomes from a gumbo of undifferentiated DNA, researchers could now pick out a single cell of interest in a microscope, isolate it, and sequence its entire genome.2 The process itself had been invented in 2002, and by 2012 it was fully automated. Sample water flowed through a thin capillary, sensors detected the presence of a cell, and some hapless microbe would then be shunted off for sequencing and then immortality in a genetic database.

But in some sense, these organisms were still digital ghosts. All we knew about them was their genetic code (sometimes just ribosomal sequences, sometimes whole genomes, more often fragments). But what were their hopes, dreams? What did they like to eat? Did they prefer methane to hydrogen? Would they rather swim towards the light or away from it? None of this leapt out of the gene data. And there were foundational questions to grapple with, too. If a scientist fully sequenced the genome of an organism she had never seen, could she be said to have discovered it? Was it okay to give it a name? Was microbiology even a laboratory science anymore, or would everyone end up sitting behind computers, looking for matches in a sequence database?

The mycologists took all this especially hard. Their domain of study, fungi, was confronted by the same crisis of overabundance as microbiology. Mushroom hunting had always been a genteel science where you were expected to produce a ‘type specimen’ that would then be the official embodiment of your discovery. You were supposed to discover fungi by walking around in tweed and lifting leaves with your cane. The new molecular techniques confronted the hunters with a flood of sequenced genomes for single-celled fungi. The four familiar, cherished phyla everyone had grown up with soon expanded to twelve. There were an awful lot of unsuspected mushrooms growing out there in the dark.

Virologists, more accustomed to being pantsed by Mother Nature, took the explosion of variety in their field more philosophically. But they struggled with definitional issues that were even trickier than the ones confronting microbiology. Viral DNA was everywhere in environmental samples, with no obvious way to organize it, or associate it with a host organism. Even the boundaries between virus and non-virus were getting murky. Many took to drink.

A big lesson of this era was that you only ever find what you’re looking for. A large branch of the tree of life called the ‘Candidate Phyla Radiation’ escaped detection until 2015 because researchers were using the wrong mesh size to filter samples (no one expected bacteria could be so small), and because CPR bacteria had ribosomal DNA that was invisible to so-called ‘universal’ primers, which turned out not to be so universal.3 In the same spirit, a novel category of virus discovered in a water tower in 1992 sat unrecognized for eleven years in an archive because it was so huge that everyone mistook it for a microbe.

A consistent pattern in the new era of discovery has been that some bizarre microbe is discovered living in a preposterous environment like a hydrothermal vent, and then a couple of years later we find out it also makes up 40% of our intestinal flora. For example, Saccharibacteria are ubiquitous in the oral cavity, but the branch of life they belong to was only discovered in 1996 (in a German peat bog), and it took until 2015 to culture the first representative of the phylum, an absolutely tiny fellow who lives on the back of bigger, tougher microbes that like to hang out near the gum line.

Surprises like this were everywhere. Equally surprising was the aggressive ubiquity of life. It became difficult to find places on Earth that were uninhabited. You could collect and sequence DNA from cloud tops, hypersaline lakes, or the coolant loop in nuclear reactor cores. Drill kilometers under the sea floor and you would find living bacteria that perked up when you fed them, after a 100-million year nap.

Another key lesson was that microbial communities formed a more natural unit of study than individual species. The reason many of those Candidate Phyla Radiation beasties were so small is because they were missing some basic machinery of life, and lived as obligate symbionts (or parasites) of larger hosts. The same was true for the tiny class of archaea called DPANN. Many of these microbes lived off of one another’s waste products and borrowed molecular machinery as needed, the way a neighbor might borrow a cup of sugar. Some organisms could swap out entire metabolic pathways depending on what food was available at the moment.

Microbes also never took high school biology class, and so were unaware that genetic inheritance can only be passed down the germ line. Instead, they traded genes between themselves like Pokémon cards, in a promiscuous process called horizontal gene transfer. The microbial tree of life had a lot of cross-links. And a bewildering menagerie of viruses, viroids, phages, and other entities complicated the matter further by smearing genetic material across species, phylum, and sometimes even kingdom barriers.

By 2016, we had a beautiful new tree of life. The red dots in the diagram below show lineages that are known from environmental DNA, but do not have an isolated representative species.

For biologists, this period was a little bit like what astronomers went through in 1923, after discovering that stellar nebulae were entire galaxies in their own right. Imagine going outside to find that the Louisiana Purchase has been added to your backyard. There was a vertiginous sense of horizons expanding, and a glimpse of the full scope of human ignorance, which is rarely granted to a scientific field.

It all prompted a natural question—just how much more of this stuff was out there to find? In 2016, Kenneth Locey and Jay Lennon argued from mathematical models that we should expect to find a trillion species of microbe. Their claim started a debate that ranges across six orders of magnitude, from skeptics who argue for millions, through a middle ground of hundreds of millions, up to the enormous original claim. Whatever the answer is, it dwarfs the 30,000 known microbes that have so far been cultured.

This leaves the field in a bind. There is both too much and too little data at the same time. On the one hand, microbiology is drowning in sequences; the challenge is to find organizing principles and concepts that might make the great wealth of genetic data more tractable to human understanding. At the same time, we barely know anything about the microbial world, and need to keep filling gaps in the data with systematic surveys of the human microbiome and every kind of terrestrial niche.

So the answer to the question I posed at the start of this essay is that it’s complicated.

We can reliably find life if we know exactly what we’re looking for. But it is not true that, given a microbe chosen at random, we could identify it even at the superphylum level. Most of the weird DNA we find in environmental surveys gets binned in a category called ‘other’ that remains opaque to our understanding. We can’t even compile a definitive list of the microbes that live in the human bloodstream, let alone the species that might journey with explorers on a spacecraft.

Nor is it true that microbes are little bags of genes that stay the same over time, and that these genes uniquely determine an organism’s position in the great tree of life. Microbes exchange genetic material at the drop of a hat, a process that makes categorizing them in the wild frustrating, and one that has dire implications for any contamination scenario where terrestrial organisms encounter distant living relatives on Venus or Mars.

Let me talk about these implications of all this for space exploration in a little more detail.

Implications for Planetary Exploration

1. The biggest lesson of this new era in microbiology is that we’re not very good at finding life. Microbes that have been living not just under our noses, but physically inside our noses, remain unknown to science. The most prevalent organism on Earth, a free-swimming ocean microbe called Pelagibacter ubique, was only identified in 1990, and not successfully cultured until 2002. A class of bizarre microbes that make hydrazine (rocket fuel) as a metabolic intermediate turned out to be responsible for half the ocean’s nitrogen production, even though scientists until the 1990’s were skeptical that their brand of metabolism was even possible. These organisms, the annamox bacteria, were not identified until 1999. The list of these things is endless.

   And this is just the state of things on Earth, a planet with known biochemistry, big comfy labs, and armies of grad students. You can imagine the difficulty of trying to distinguish native from Earth life with the kind of rudimentary equipment we would send to Mars or the cloud tops of Venus.

Read more

After two years, four prototypes, and a great deal of paperwork, Lavazza and the Italian space agency sent a proper espresso machine to the ISS in 2015. On Earth, a basic Lavazza espresso maker costs about $150 and weighs 3.5 kilograms. The coffee machine’s spaceborne cousin was a 20kg box about the size of an oven. The cost to build it was not disclosed, but was likely in the single-digit millions

Asking how a coffee machine got to be so huge and expensive in space is a good way of understanding the cost drivers in human space flight.

Espresso machines are not particularly lethal on Earth, but almost anything on the space station can kill the crew if it’s built wrong. So the ISSpresso had to prove to NASA’s satisfaction that it would not take out the station’s electrical system, interfere with the radio, leak boiling water, catch fire, dazzle the crew with bright lights, electrocute anyone, be dangerously hot, make loud noises, emit noxious gas, shatter into fragments, smell weird, or shake apart in the harsh conditions at launch. (The sharp pin that punctures the coffee capsule required a special safety waiver.)

The authors of the technical paper on ISSpresso include a list of some of the NASA standards they had to comply with to get their machine certified for launch and orbital coffeemaking.

These documents are not light reading. It can be tempting to dismiss them as NASA run wild, and there are certainly some requirements (like handle shape or enclosure color) that seem arbitrary. There is also a lot of bureaucratic connective tissue, like the standards for harmonizing processes between NASA and the European and Japanese space agencies, who all build their hardware to slightly different specs.

But most of the technical requirements in this list have substance. They fall into a few broad categories:

• Making sure nothing on the payload damages the space station, either in normal operation or if something goes haywire.

• Lots and lots of fussiness about electrical behavior and electromagnetic interference.

• Consistency in interface design with other ISS hardware.

• Demanding proof that the ISSpresso can take a physical beating (especially during launch), endure kicks from astronaut feet, sudden decompression to vacuum, abrupt surges or sags in voltage and water pressure, and other environmental insults.

• Ensuring the heating element doesn’t burn anything or set itself on fire. This is a trickier requirement in space, where air doesn’t cool things by convection.

• Fluid handling requirements specific to the zero g environment. The ISSpresso has to contain spills and not fill the cabin with a mist of boiling water. It also has to play nice with ISS plumbing.

• Astronaut-proofing the enclosure, which will inevitably be kicked and used as a handhold. This includes making sure nothing can hurt a clumsy astronaut (sharp edges, pointy switches, pinch points) or break if it’s yanked on.

• Proving that the ISSpresso won’t shake apart during launch or damage whatever it launches with.

• Antimicrobial measures for all wetted surfaces and plumbing.

• Basic OSHA-type stuff like noise limits.

None of these requirements are frivolous, and some of them reflect dangers unique to spaceflight. If the plastic cover shatters on your espresso maker back home, you’ll be mildly inconvenienced. But if that cover shatters in space, it can pose an acute inhalation and eye hazard.

The many technical requirements are enforced by the Safety Review Process, itself a highly regimented standard. The Process takes designers through a series of project milestones and official reviews that ultimately satisfy NASA that each requirement on their lists has been met.

The Safety Review Process begins with a friendly chat about general design ideas, and then ratchets up in rigor and unpleasantness. By the final milestone, a NASA bureaucrat is shining a light bulb in your face and screaming at you to confess everything you know about mission risk. It’s not enough to tell NASA that you plan to put your payload on a truck and drive it to Kennedy Space Center for launch; you have to analyze the g-forces for every crane movement and specify how fast the truck will go. Any conceivable failure mode has to be identified in a Hazard Report, along with the proposed fix, and that fix has to be certified.

There is a truism in aerospace: when you pay $500 for an aviation-certified thumbtack, what you’re really paying for is the ten binders of compliance documents, certifications, and tests that accompany it through the production process, along with a promise that someone will go to jail if any part of that process is falsified.

The Process is painful, but it’s not unique to NASA. We run versions of it in aviation, military, and medical contexts, wherever human lives are at stake. It is often ridiculous and everyone hates it. But some version of it is the only way to be sure systems behave as intended.

Let me illustrate this with a moving personal anecdote!

A Moving Personal Anecdote

I live in a solar-powered home in rural New Mexico. The house is not connected to the electrical grid; instead, power from solar panels feeds a rack of batteries, and a machine called an inverter draws power from the batteries and turns it into household current.

The solar system in my home is supposed to be decoupled. One wall of the electrical closet has all the solar gear; the other has a standard junction box with circuit breakers like you find in a normal home. From the house’s perspective, alternating current flows in just like it would from a power line. And on the solar side of the system, the inverter doesn’t know or care about what’s happening inside the house. As long as the total power draw stays under a generous maximum, everything is supposed to just work.

That’s the theory.

But after upgrading the inverter last year, I found myself beset by electrical gremlins. A few times a day the lights would dim, and I could hear the pump in my aquarium start to make a choking noise. At those times, a display on the inverter showed the A/C voltage dipping. Sometimes the inverter would reboot, taking down power for the whole house for a minute. There was no discernible pattern in when or how often this happened.

I thought I could live with the problem until it started killing my furnace. The first couple of times, the victim was a transformer, a $25 part on the circuit board that I learned to replace myself. But the third time around, the voltage drop burned out the entire logic board, forcing an expensive repair that left me without heat for a week.

At this point it was November, and heating the house had become a game of Russian roulette. I knew that every minute the furnace stayed on, a blip in the electrical system might kill it. No one I talked to could identify a cause. I had to figure out what was causing the drops in voltage before the house became unlivable.

Being a software guy, I decided to try binary search. I turned off half the circuit breakers to the house one day, then the other half the next, to see which side the problem was on. Soon I had isolated it to one part of the house, and then to a single circuit in the bathroom.

There I found the culprit: a Japanese shower toilet.

The toilet had a small heating element that turned on and off to keep the water in the bidet attachment and seat warm. Whenever the heater came on, its modest appetite for electricity was somehow enough to destabilize the inverter, which then briefly delivered lower voltage to the entire rest of the house. While most appliances could handle these dips, the furnace could not, and died dramatically. Even though the toilet’s power demand was low, there was something about its Japanese expectations for voltage and frequency (just a little bit off the US standard) that made the American-made inverter crazy.

Figuring that out took me several weeks and a few thousand dollars. My mistake was believing that the power system really was decoupled—that nothing in the house could affect things upstream of the junction box. That is what the inverter specs and circuit diagrams all said. That is what customer support told me. But it wasn’t true.

Since that time, I’ve learned that small heaters (like coffee makers or kettles) can be kryptonite to an inverter, and that this is common folk knowledge among solar installers. But the consequence, that a guest can do damage to my home by plugging in a hair dryer, is still unsettling and counterintuitive.

This is the class of problem all those NASA interface requirements are trying to forestall. If you’ve ever had a faulty wiring harness in your car (hello Jeep owners!) you know what a nightmare it is to try to chase down intermittent, poorly localized faults. NASA inflicts eye-watering certification costs on itself and its partners to avoid trying to diagnose this stuff in space, where half the systems can’t be powered off, and where there’s a high chance of killing the crew if you break something.

Undoubtedly, some proportion of NASA’s Safety Review Process is overkill. But even if we could cut regulatory overhead by 75%, a device like the ISSpresso would still cost a few hundred thousand dollars to develop and end up built like a tank. The blast radius of malfunctioning hardware on human-rated spacecraft is simply too big to avoid doing some version of the safety dance.

This has uncomfortable consequences for space dreamers.

There is a widespread belief that launch costs are what has been holding back space exploration, and a corresponding excitement now that they are dropping by a potential two orders of magnitude. Many SpaceX fans in particular believe that Starship solves every problem by being huge and cheap. And they are partially right! It would be much easier to send people to Mars on a 1200 ton rocket than to try to fit all the equipment they need into a 60 ton transit habitat engineered like a Swiss watch.1

But cheap launches can’t solve the equipment problem. Ultimately, whatever we put inside the spacecraft has to work as advertised, and until we have hundreds of person-years of experience living in space habitats, the only way to guarantee that will be an expensive process of flight qualification and testing.

That means future human missions to space will have the same cost profile as big space telescopes do today—a few hundred million spent to launch stuff, and billions spent inventing equipment and trying to get it to work right.

No Espresso on Mars

Like all our problems, this one gets worse on Mars.

The defining feature of a human mission to Mars is that risks are sequential and cumulative. Every link in the chain has to go right, or the mission fails. This means early visits to Mars will have safety and reliability requirements that make the Space Station look like a middle school science fair.

These requirements will be especially tight for the surface part of a mission. Any equipment that lands on Mars will have to demonstrate that it can launch from Earth, sit dormant for six months, survive entry and landing, and then work in partial gravity and dust without breaking for 17 months. Machinery that is pre-positioned on Mars in advance of the crew (a common risk-cutting measure in mission designs) will also have to prove that it can sit out in the weather for two or more years.

To make matters worse, any payloads sent to the surface will be severely constrained by weight. This is not for want of big rockets to send them to Mars, but a consequence of the fact that landing heavy payloads is hard, with the difficulty going up as some integer exponent of the landed mass.2

Whenever you need a combination of light weight, reliability, and autonomy in a space context, it is time to bring your wallet to your lips and kiss it goodbye. We saw an example of this last week in the context of Mars Sample Return, where a rover whose sole purpose was to move a few titanium tubes from the ground into a box wound up costing half a billion dollars.

The same pathology is going to bedevil us when we finally get to Mars, even if launches there from Earth are free.

What is to be done?

It’s pretty frustrating to enter an era of cheap rockets and still not be able to do fun things. Early Mars concepts (like NASA’s Design Reference Architectures) agonized over how to fit the mission into the minimum number of launches, which were the most expensive line item in the budget.

If Starship and New Glenn succeed, we can have all the mission mass we want. But that just runs us into the next-biggest item on the cost list, the reliability and testing issues that are the subject of this post.

So what do we do to make certification and testing cheaper?

1. Fly more. If there are a dozen space stations that all need an espresso maker, then that makes designing ISSpresso 2.0 and later models much easier. A proven flight record replaces a lot of first-principles testing.

2. Fly more robots. Robots don’t drink coffee, but there are science missions that could use a pressurized hot water source, and validating such equipment where it doesn’t pose risk to astronauts makes it easier to adapt it for human space flight later. This holds for all kind of devices and sensors that would be useful on manned spacecraft.

3. Learn to land on Mars. Right now we can land 2-3 tons on Mars at a time, in an error ellipse that is about 20 kilometers long. For a realistic human mission, we need to be able to land 100 tons or more at 100-meter precision, so that we can pre-position equipment and land in our favorite crater. This capability would also make it cheap to send big dumb robots in large numbers to Mars, instead of the very expensive, artisanally hand-crafted robots we send now.

4. Fix the safety ratchet. It is easy to add safety constraints and hard to undo them. You and I will probably die before we’re allowed to take a bottle of water through airport security again.

   Many NASA rules around software reliability date back to the 1970s and don’t make sense in the smartphone era. Harsh limits on electromagnetic interference impose a testing burden on innocent components that probably don’t need it. And some of restrictions on flammability and wiring are a hangover from the Apollo I fire in 1967. There needs to be a mechanism for relaxing rules to adapt to changing conditions, or else the space program will fossilize in its own paperwork.

5. Let amateurs fly stuff. People are inventive, and we should let gifted engineers try things in space without interference from the safety bureaucracy, as long as they don’t hurt anyone. Hopefully the new era of cheap launches will enable some risk-taking and invention by talented amateurs, and the stuff that doesn’t blow up can then carry over into our official space program.

   Mars For The Rest of Us is fantastic! Please subscribe today for more, more, more.

Further Reading

1. The technical paper on designing the ISSpresso is very readable and fun. ISSpresso Development and Operations (2015) DOI 10.1016/S2468-8967(16)30038-6

2. For more on the special challenges of handling liquids in space, along with the vaguely vaginal coffee cup invented for space use, see How Advances in Low-g Plumbing Enable Space Exploration (2022) DOI 10.1038/s41526-022-00201-y

3. Behold in all its splendor the Pressurized Payloads Interface Requirements document for ISS.

So another idea is to bring rocks back home. And so sample return has been a top goal of planetary science for thirty years or more.

There have been some successes! Most famously, the Apollo astronauts brought back a few hundred kilograms of moon rocks. In 2004, a mission called Stardust flew past the comet Wild 2 and brought back dust grains from the comet’s coma. In 2005, the Japanese Hayabusa probe landed on an asteroid called 25143 Itokawa and brought back a little surface dust for study. Its successor, Hayabusa2, did the same for an asteroid called 162173 Ryugu in 2018. And in 2023, a mission called OSIRIS-REx returned to Earth carrying a sample from the asteroid 101955 Bennu.

Investigations from these missions have been fruitful and have built expectations for a Mars sample return mission, which would bring the full power of Earth laboratories to bear on some astrobiologically interesting rocks the Perseverance rover found on Mars. The circumstantial evidence for early life on Mars is now strong, and there’s a widespread hope that analyzing samples on Earth would have spectacular results, more than enough to galvanize public support for future Mars missions.

The trouble with grabbing samples from Mars is that the planet has more gravity than any asteroid or comet. To lift a 10 kg sample off the surface to orbit takes a 350 kilogram rocket. That 350 kilogram rocket has to land on a specialized vehicle weighing a couple of tons. The lander in turn has to enter the Martian atmosphere in a suitably large aeroshell, which then needs a rocket big enough to launch it in one piece from Earth. These kinds of requirements propagate across the mission, and pretty soon you’re talking real money.

NASA also holds some deeply fundamentalist beliefs about planetary contamination that complicate the mission and (crucially) increase the weight of any return container. Comets and asteroids are considered low-risk and don’t require this level of fussiness, but NASA treats Mars as if it’s knee-deep in space Ebola, creating a lot of hoops for itself to jump through to prevent the smallest speck of frozen Mars dust from killing us all.

All of this makes the sample return mission NASA and the European Space Agency embarked on in 2018 the most complex and ambitious planetary mission design ever attempted, with multiple spacecraft and rovers passing samples to one another in (depending on your point of view) either a graceful cosmic ballet, or a Rube Goldberg design from an agency that has lost the ability to do things simply.

The easiest way to describe Mars Sample Return is to introduce the players.

The Perseverance rover, which landed in 2020, is the one part of the Mars sample return program that made it to Mars. The rover has been collecting samples around Jezero crater and storing them in a set of 43 titanium tubes designed to be returned to Earth.

Before leaving Jezero in 2022, Perseverance left ten tubes cached at the bottom of the crater, as an insurance policy. The rest of the samples remain on board the rover, and will either stay there or be dropped near a potential landing site, depending on the state of health of the vehicle. NASA recently certified the plutonium-powered rover for another 100 km of travel and expects it to remain active into the 2030s.

The Orbiting Sample is a small 5 kilogram basket that resembles a salad shooter. It is designed to bring the sample tubes collected by Perseverance back to Earth, and forms the through line for the sample return mission—its weight and dimensions set the size of pretty much every other spacecraft in this list.

The Orbiting Sample is meant to arrive on Mars on a lander, which also carries a rocket to fire it back into space. Once the sample tubes are secured inside, the rocket launches the assembly into low Mars orbit, where another spacecraft is supposed to find and collect it, performing some fussy cleaning steps during the process. This second spacecraft then puts the Orbiting Sample inside a heat shield and returns it to Earth to crash-land gently in the Utah desert.1

At that point there are more fussy cleaning steps before the recovered assembly is taken to a special containment facility to begin the long work of unboxing samples and distributing them to specialized laboratories for study.

The Mars Ascent Vehicle is a stubby little rocket whose job it is to put the Orbiting Sample into Martian orbit. The rocket has to be powerful enough to get its payload into space, but also small enough (both in terms of dimensions and weight) to fit into the lander that brings it to the surface of Mars.

One challenge for the ascent rocket is temperature. The U.S. arsenal has plenty of stubby rockets that can sit in storage for years and still fire reliably, but none of them are designed to work in conditions as cold as the Ascent Vehicle would experience on Mars. And in fact, no one has ever launched a rocket from the surface of another planet, making the Ascent Vehicle the technically riskiest link in the chain of events meant to carry the collected samples home.

The Sample Retrieval Lander is a lander of size meant to deliver the ascent rocket and empty Orbiting Sample to the Martian surface. It is supposed to land as close as it can to Perseverance (or wherever the rover finally decides to lay down its burden of sample tubes), and this need for a precision landing drives its enormous bulk. In the initial design, this big boy carried yet another spacecraft:

The Fetch Rover is a small rover with a long arm. It arrives on Mars along with the Sample Retrieval Lander, or possibly on its own (one of many parts of the design that never stabilized). The Fetch Rover’s job is to physically pick up sample tubes from the Martian surface and place them in the Orbiting Sample. In the best case, the Sample Retrieval Lander will have landed very close to the samples, and the Fetch Rover won’t have much work to do. In other scenarios, there might be some distance to cover between the lander and the sample cache, and the Fetch Rover then shuttles back and forth between them.

The final piece of the puzzle is a giant spacecraft called the Earth Return Orbiter. Its job is to find the Orbiting Sample in orbit around Mars, collect it, and take it back to Earth. The collection step is a convoluted sequence that includes zapping the exterior of the Orbiting Sample with ultraviolet light to eliminate the remote possibility of contaminating Earth with dust-borne Martian life.2 For the same reason, the orbiter diverts from Earth into a lonely graveyard orbit around the sun after delivering its payload.

And that, in its essentials, is Mars Sample Return. For those keeping track, the mission includes two rovers, two orbiters, three launches from Earth, one first-time-ever launch from Mars, and a challenging treasure hunt in low Mars orbit for the Orbiting Sample, which carries no beacon and is about the size of a basketball.3 Two of the vehicles needed—the Earth Return Orbiter and the Sample Return Lander—would be the largest spacecraft of their kind ever built.

If this design strikes you as a little much, you are not alone. The only mission remotely comparable to Mars Sample Return is the original plan for the Artemis moon landings, which included double-digit launches and a complicated cast of spacecraft (Orion, Gateway, tankers, the Human Landing System) performing an orbital dance that left the nagging impression that there had to be an easier way to do all this.

There are some oddities to the mission architecture. Perseverance was designed so there is no way to remove samples from its interior if it dies—NASA would have to make a decision to drop them while the rover was still working. The Fetch Lander cost more than half a billion dollars and by design did almost nothing. Its job was to hand samples to the lander, which could also receive them directly from Perseverance. The Earth Return Orbiter would be, for no clear reason, the largest space probe ever built, powered by a suite of solar panels more than forty meters across. The Sample Return Lander weighs three times more than anything anyone has ever tried to land on Mars before.

The sole purpose of this beefy team of robots was to return about 500 grams of material from Mars to Earth. But as the mission blew through its budget estimates and started looking for things to cut, the inevitable happened. NASA started reducing the number of samples the return mission would carry. Congress, lacking an appreciation for the absurd, killed the program before NASA could take the process to its logical conclusion and design a sample return mission that would come back to Earth carrying nothing. But the result was much the same.

It’s notable that Chinese scientists, who are flying their own sample return mission to Mars in 2030, have chosen a design with just two spacecraft, a sampler/ascent rocket and orbiting return vehicle. Critics of the Chinese plan argue that this ‘grab and go’ sample mission has less nuance and scientific value than the carefully curated suite of samples waiting on board Perseverance. The Chinese would probably counter that much of the science value in sample return goes away if you don’t actually return the samples to Earth.

A Brief Mission Timeline

The reasons Mars Sample Return died are interesting and complex enough to deserve their own post. For now I’ll start with a timeline of the project, to help readers keep track of its various twists and turns later.

April 2018 - NASA and the European Space Agency (ESA) agree on a tentative sample return mission to fly in 2020.

July 2019 - NASA and ESA settle on the basic architecture of the mission. ESA will build the Fetch Rover and Earth Return Orbiter, NASA carries the rest.

July 2020 - Perseverance departs for Mars on an Atlas V rocket.

October 2020 - The first Independent Review Board report on Mars sample return makes numerous recommendations while continuing to strongly endorse the project. The estimated development cost to NASA at this point is $3.8-$4.4 billion.

February 2021 - Perseverance lands in Jezero Crater and begins roving.

April 2022 - The National Academies release the Planetary Science Decadal Survey, a consensus document that sets the scientific community’s official priorities for the next decade of planetary exploration. As in the previous two decades, the survey reconfirms Mars sample return as the top priority for planetary science.

July 2022 - The Fetch Rover is cut from the mission design. Perseverance is now expected to deposit the samples directly into the Return Lander, with a pair of Ingenuity-style helicopters (!) added to the mission as backup.

May 2023 - NASA establishes a second Independent Review Board, making sample return the first NASA science mission to require two review boards.

September 2023 - the new IRB report says NASA has no chance of meeting its stated budget or timeline. The report estimates sample return will cost up to $11 billion and might not return a sample to Earth before 2040.

November 2023 - the program is put on hold while NASA contemplates how to respond to the IRB report.

April 2024 - NASA Administrator Bill Nelson calls both the cost and timeline of the mission unacceptable, and asks industry to submit proposals for flying the Mars sample return mission more cheaply.

June 2024 - NASA selects eight outside proposals (from about forty received) for further study, plus three proposals from inside the agency.

July 2024 - A paper announces that Curiosity has found strong evidence of ancient microbial life in an ancient lakebed deposit at Gale Crater.

January 2025 - NASA announces it will explore two landing options for Mars Sample Return. The first is basically the status quo, the second “will capitalize on using new commercial capabilities to deliver the lander payload to the surface of Mars,” a phrase most observers interpret as a reference to SpaceX’s Starship rocket.

April 2025 - NASA publishes a revised architecture that uses a smaller ascent rocket and moves the cumbersome sterilization step from the Earth Return Orbiter to the surface of Mars.

May 2025 - Trump’s proposed budget cancels Mars Sample Return, part of a wider gutting of science programs at NASA.

September 2025 - Imagery and sampling from Perseverance show strong evidence for ancient microbial life in a mudstone at Bright Angel, near Jezero Crater.

January 2026 - Congress passes a NASA funding bill with no provision for Mars Sample Return beyond a $110 million line item marked ‘Mars Future Missions’.

February 2026 - Musk announces SpaceX is putting its Mars plans on hold to pivot to the Moon, eliminating Starship as a viable option for saving the mission.

There’s not a big difference in energy between the two destinations. Any rocket that can land on the moon can easily put a crew in Martian orbit. The issue is time.

As we saw recently with Artemis II, a spaceship can get human cargo to the Moon and back in about ten days. But orbital mechanics makes it hard to complete a trip to Mars in less than two years, and rigid launch windows further constrain options for abort or rescue. Bridging the gap between the two weeks we’ve spent on the Moon and the long, committal journey we’d have to make to Mars runs us into a thicket of difficulties.

In other words, the ladder to the stars is missing some rungs.

It would be nice if there was a class of mission intermediate in difficulty between the Moon and Mars, one that didn’t take us so far out of our experience base and had better abort options than ‘press the red button and wait two years’. Even better if the mission had a milder radiation environment, shorter communications delay, and a high potential for scientific discovery.

This class of mission exists, but no one likes to talk about it:

An orbital trip to Venus is like one of those gorgeous high-rise apartments in Jersey City. Everything about it is perfect except the location.

And I understand the reluctance! Venus is the biggest heartbreak in the Solar System. The planet could have been Earth’s twin, but instead became an acid-washed nightmare and climatological horror story. People have never recovered from the big reveal in the early Space Age that the clouds enveloping our beautiful planetary neighbor covered a hell world.

But with Venus, you have to adopt a ‘glass half full’ approach, even if the thing the glass is half full of is concentrated sulfuric acid.

To get the bad stuff out of the way, the surface temperature on Venus is around 470°C, about as hot as a pizza oven, and the pressure is 92 times what we enjoy here on Earth. The high cloud layers, while temperate, are mostly sulfuric acid. And because volcanic activity has reworked the planet’s surface within the last billion years, there is little hope of finding Mars-like relics of the planet’s habitable past, even if we could build the rovers to look for them. If young Venus had temperate oceans or harbored life, the evidence for it has been buried about as thoroughly as anything in this solar system can be buried.

So that’s the bad part. But once you move past it, you start to notice that everything gets easier on Venus. The atmosphere is great at blocking radiation. Solar panels can be small, and the pressure and temperature in the high clouds are so Earthlike that, if not for the acid, an astronaut could sit in the gondola of a Venusian blimp wearing only an oxygen mask and a swimsuit.

Launch windows to the planet recur every 19 months, compared to 26 months for Mars, and the round trip communications delay is about half of what a Mars-bound crew will face. Abort trajectories still aren’t great, but are about twice as fast as trying to make it home from Mars. And everything goes faster: the orbital mission I cited would get the crew home just three days past the absolute duration record for human spaceflight.

Even the gravity on Venus (0.91g) is homelike, which means that airship habitats, sensors, smoke detectors, toilets, and all the rest can be developed on Earth instead of forcing us to build a space station that can simulate Martian gravity. An entire class of hard problems around physical deconditioning just goes away. Astronauts flying the friendly Venusian skies could recover from the short transit from Earth in a sunny, homelike environment, and if they ever wanted to roast a turkey, they would only have to lower it a few kilometers through the atmosphere on a fishing pole.

The Science Case for Venus

There are anomalies in the Venusian atmosphere that are consistent with the presence of life. Unlike on Mars, where any living microbes are presumably deep underground, life on Venus can only exist in the clouds and should be easy to observe (or rule out) with cheap balloons and aerostats.1

Maybe the strongest piece of circumstantial evidence is the detection of phosphine, a gas that has no business being in the atmosphere and should degrade rapidly (within days or weeks) in sunlight. The phosphine detection was controversial when it was first announced in 2022, but it has since been corroborated by multiple measurements. Like the detection of methane on Mars, the pattern of occurrence of the gas remains puzzling.

But the presence of phosphine is just part of a larger pattern of anomalies:

• The lower part of the Venusian atmosphere contains an ‘unknown absorber’ that captures about half of the incident ultraviolet light. Its changing large-scale patterns and absorption characteristics are reminiscent of phytoplankton blooms on Earth.

• There is an unexpectedly high proportion of water vapor to sulfur dioxide in the high atmosphere. The two gases share a common volcanic origin and should have a similar abundance profile, but SO₂ is significantly depleted.

• Ammonia exists in the cloud layers, although it should not be possible for it to be there.

• There is molecular oxygen in the clouds with no identifiable source.

• The clouds at around the 45 km level contain a population of particles called the ‘Mode 3 haze’ . The composition of this haze is unknown, but the particles that comprise it are not spherical, and so cannot be droplets of liquid.

• The atmosphere in general is not in chemical equilibrium.

I wrote a few weeks ago that the key constraints for life on Earth are temperature and water activity. For all that the climate in the Venusian clouds is balmy, it’s insanely dry, a hundred times drier than the Atacama Desert. Any life that existed there would have to fight for each molecule of water against sulfuric acid, which is desperately hygroscopic. While complex organic chemistry can exist in acid droplets, it would not resemble anything like the biochemistry we know on Earth.

This is why the detection of ammonia in the otherwise hyperacidic atmosphere is so intriguing. If there’s a mechanism that raises the pH in at least some droplets to around ~1, then Earth-like extremophiles could thrive and be able to photosynthesize in the clouds of Venus2.

Venus researchers Janusz Petkowski and Sara Seager have made an intriguing case that all the anomalies I list can be explained by the presence of microorganisms that produce ammonia from nitrogen and water. The ammonia would react with sulfuric acid to create a kind of slurry, accounting for both the depletion of SO₂ and the Mode 3 haze. The pH in such a neutralized droplet would be close to 1, a perfectly livable environment for acidophiles we find on Earth. And a byproduct of their metabolism would be molecular oxygen.

The appeal of this theory is that it’s ridiculously easy to test. Unlike Mars, where we have to delve deep underground to hope to find relic life, we can just go look at the clouds on Venus with a party balloon. The mission is simple enough that Petkowski and Seager are flying a private version of it with RocketLab as a side project, funded by an anonymous benefactor.

The way I like to think about this question is that we can’t lose. Missions to the clouds of Venus are either going to find life or some kind of brand new chemistry, either of which will be a breakthrough discovery in planetary science. There’s basically a guaranteed Nobel prize waiting in the skies of Venus for whoever wants to collect it.

A more sober case for exploring the planet is that we only have three terrestrial worlds to work with. We should learn all we can about how they formed, how they function, and why their fates diverged if we want to better understand exoplanets that humanity won’t be able to physically visit for millennia. Right there is no way to distinguish Earth-like from Venus-like worlds in exoplanet surveys, or even to identify meteorites on Earth that have a Venusian origin. What we learn on Venus will compound our understanding of planets across the sky, not to mention our own planet’s climate.

As a NASA paper points out, humanity has accumulated 31 years of surface time on Mars and 49 years in Martian orbit, but we’ve spent just 4.5 days spent exploring the Venusian atmosphere, and 9.4 hours on the surface. When you consider that every mission to Mars has brought fundamental shifts in our understanding of the solar system, it would be strange if Venus didn’t have surprises waiting for us as well.

Exploring Venus, from Easy to Hard

There is no planet friendlier to the extraterrestrial balloonist than Venus. Since the gravity, pressure, and temperature in the high atmosphere are all close to Earth, all you need is to apply a little acid protection to your aerostat, and you’re good to go.

And in fact, balloons on Venus were the first aircraft to fly off Earth. Two Soviet Vega probes in 1985 each delivered a French-made helium balloon to the atmosphere. The balloons cruised around at 54 kilometers for about two days before their batteries ran down, covering about 10,000 kilometers in the process.

There are different balloon designs, from easy to hard. The easiest type to fly is a fixed-altitude balloon, which is just a weatherproofed version of the Chinese spy balloon that terrorized America back in 2024. Such a balloon would drift around with the wind and could take advantage of the fact that solar panels in the reflective clouds can point in any direction, with no need to track the Sun.

The next step up from this design is a variable altitude balloon. These balloons use pumps, reservoirs of boiling/condensing water, or just brute force to compress the lifting gas in their envelope, allowing them to vary their altitude by several kilometers. Such a balloon could dip down into the lower cloud layers that are too hot for sustained operations, as well as explore conditions through the full putative habitable zone (45-65 km).

The most ambitious balloon design is a hybrid balloon/flying wing that looks like a B2 bomber that really let itself go. The flying wing would be buoyant, but also sufficiently plane-like to move freely around the atmosphere. The hard part about getting one to Venus is the atmospheric entry. The behemoth would have to serve as its own heat shield, which means it would probably take several practice attempts to stick the initial landing.

And finally, there is the solar-powered airplane. A large enough model could outfly the winds and linger on the sunlit side of the planet for weeks or months. The difficulty here again is delivery. A plane would have to fold up tightly to fit into an aeroshell, and how to get it unfolded and flying in midair is the big technical challenge.

But these are fun problems to have! The science return on any airship design with 2026 sensor technology would be phenomenal, and they could all be rigged to drop a series of sondes or mini-landers down to the surface.

Surface Exploration

There are three approaches to handling the problem of heat on the surface of Venus.

The simplest one is denial. You send down a chilled probe wrapped in all the insulation you can manage, and try to arrange a high-bandwidth conversation with an orbiter to maximize the amount of data the lander sends before it cooks. The Soviet Venera landers, the only spacecraft ever to take pictures of the surface, followed this approach. They kept their instruments in near-vacuum inside a titanium sphere and likely survived for several hours after landing (their useful transmission time was limited by line of sight to the orbiter).

A second approach is to put the more sensitive electronics in a refrigerated box. Refrigeration on Venus is a relative term—the cold box on a chilled lander will still be at around 200°C, enough to beautifully roast a chicken. And running a fridge on Venus means having a hot-side radiator at some frightening temperature like 1200°C. But the overall concept is simple: you use electrical power to keep an insulated compartment significantly below ambient temperature, and maybe put in some sapphire windows for the cameras to look through. Such a hybrid design extends the lifetime of a lander from hours to several days or more.

The final and most metal approach is to dispense with refrigeration entirely. NASA has been experimenting with integrated circuits made from silicon carbide that can take a thermal beating. The Glenn research lab has kept chips running at temperatures over 500°C for a year, and even built prototypes that function at 900°C. These electronics are primitive, but more than capable of handling signal processing, amplification, basic imaging, and many of the other tasks you want in a Venus lander. There are even motors strong enough to turn the wheels on a rover. The technology level of this stuff is early 1970’s, but when you consider what kind of data we got from Viking and Voyager, the prospect of doing sustained red-hot science on the surface is thrilling.

Powering a long-duration surface probe is a challenge. Because the surface is so hot and dim, solar panels are impractical. One study estimates solar panels on the surface of Venus would generate about 8 watts per square meter, compared to ~500 watts at an altitude of 60 kilometers, and ~2000 watts in low orbit.3 Industrial batteries exist that work at temperatures near 400°C, but there is still some work to be done to get them functioning at the ambient temperatures on Venus. Nuclear RTGs could work, but have to run extremely hot.

On the plus side, because the atmosphere at ground level is thick, a small wind turbine might be enough to keep a battery topped up indefinitely. As long as we can live with low data rates, a next-generation lander could stay alive on the surface for several months.

Manned Missions to Venus

Read more

While capsules like Soyuz and the Crew Dragon are equipped with a rudimentary toilet kit, astronauts use a mix of drugs, diet, and occult knowledge passed down the generations to keep from having to use it. It can take more than two days for a Crew Dragon capsule to reach the space station, but the crew in the cramped spacecraft is expected to save the real fireworks for the relative comfort of the space station toilet.

The longest anyone has attempted to hold it in space is Frank Borman, on Gemini 7. Stuck in a two-man capsule the size of a phone booth, Borman was determined to get through his two-week mission without releasing the hounds. His crewmate Jim Lovell describes the outcome:

STONE: Can you give me some idea how you spent 14 days in that little, cramped quarters with somebody else without going absolutely mad?

LOVELL: It’s not easy. But you have to remember that you work so long—I was there 3 years—for a spaceflight, so I’d go up with just about anybody. And Frank and I were, you know, just this close together. And you get to know each other quite well.

STONE: Yes!

LOVELL: I mean it was something. As a matter of fact, Frank went, I think, 9 days without having to go to the bathroom. And he said, “Jim, this is it.”

LOVELL: “I said, “Frank, you only have 5 more days left to go here.”

Borman would go on to suffer miserably from spacesickness on Apollo 8, giving his fellow astronauts (including poor Lovell again) an unwitting but convincing demonstration of the inadequacies of the Apollo waste collection system, from both ends.

Everyone agrees that the sanitary conditions aboard Apollo were barbarous. Going to the Moon in the tiny capsule was like living in a three-man port-a-potty, made worse by the fact that doing the deed took the best part of an hour, with much of that time spent kneading antimicrobial powder through the contents of the collection bag.

But astronauts back then were hardy, they wanted to go to the Moon, and so they dealt with ‘escapees’ and the continual stench aboard their capsule without kvetching. And so it took twelve years to get from the first manned spaceflight to the first halfway-usable toilet.

Space Toilet Theory

The next time you find yourself in the bathroom, take a minute to consider the three ways gravity helps to make the process hygienic and easy on Earth:

1. It presses you down against the toilet with enough force that you can position your body comfortably, and maintain that position.

2. It pulls waste away from the body with vigor, easily overcoming contact forces like surface tension.

3. It sequesters body waste, first under a layer of water, then ultimately in a gravity-fed sewer pipe or septic tank. The water barrier in the toilet bowl minimizes the time waste is exposed to air and prevents odors from downstream parts of the system from reaching your nose.

In a zero-gravity environment, you have to find alternatives for all three jobs.

The first task, body positioning, turns out to vary a lot across people. Over the years NASA tried a lot of methods that didn’t work, including thigh straps and special rubber shoes with suction-cup soles. They came to realize that the best approach was to provide options (handholds, thigh bars, foot rests) and let astronauts figure the details out for themselves.

The second task, separating waste from the body, has to be done with air suction. The need to maintain a strong air stream means toilet seats can’t be too wide (four to six inches is the norm), making correct body positioning even more important. The various fans also make space toilets noisy, even by the cacophonous standards of space flight.

The third task, sequestering waste and controlling odor, is tricky. Urine can be collected in a funnel, where it gets mixed with an antimicrobial agent before being sucked into a storage tank. The state of the art for fecal collection is single-use porous bags that allow airflow but retain solids and water. These are tied off after use and placed in a collection cylinder, along with any gloves and wipes that the astronaut used for cleanup.

Each collection cylinder is designed to hold about ten bags, with astronauts adding compression plates between deposits as needed to squash things down. The fecal bags are not sterilized, so the lid of the collection cylinder has a permeable filter meant to trap odors while letting evolved gases escape to the cabin.

The process works but is imperfect. Odor control in particular remains a problem, and chronic sewer smell is one theory for why astronauts tend to chronically undereat on the space station.

The Rise of the Space Toilet

The original design for Apollo waste handling was too much for even the steeliest-eyed space man. Fred Hughes describes it in an oral history:

There was a toilet design that you sat on this—the idea with the designers—that you sat on the wall and you put all your apparatus inside this ‘device’ and then you would crack this valve, and it went to vacuum. [The astronauts] said, “Nobody’s going to do that. Pull a vacuum?” You can imagine just turning your whole body inside out in it. That was a big laugh, but they said, “Take it out, we’re not going to fly it, we’re not going to use that.”

And so the Apollo program got by with a mixture of condom-like penis sleeves and plastic bags.

When Skylab launched in 1973, it was clear it would need a proper toilet. The orbital laboratory had ambitious medical goals around collecting astronaut waste, in particular a desire to freeze individual urine samples from the three-man crew over the course of the mission. And at 84 days, the longest Skylab mission would exceed the ability of even the most iron-sphinctered astronaut to hold out. So a lot of effort went into designing a toilet and urine freezer the crew wouldn’t hate.

The final design was bold, with the toilet seat mounted vertically on one of the walls of the spacecraft, so an astronaut could defecate like Spider-Man. But it needed testing, and that meant finding people who could do their business in twenty seconds aboard an aircraft flying a zero-G parabola.

Here I will cite the official Skylab history:

Urination could be successfully simulated by mechanical devices, and a urine-collecting device was easy to test; but defecation could not be simulated. Test subjects who could perform on cue were needed. The Huntsville program office was able to find a few people with this talent, and in November 1969 two days of aircraft testing produced nine good “data points” for the fecal collector.

The Skylab toilet was a hit with the astronauts and helped inform the design of the next chapter in orbital elimination, the Space Shuttle toilet.

One piece of feedback from Skylab was that the toilet needed stronger airflow. This meant the Shuttle toilet opening had to be narrow. To practice correctly positioning their body, astronauts on Earth sat on a special training mockup with a camera mounted in the center of the waste tube. A successful docking with the device meant precisely centering one’s nether eye in the crosshairs of a video screen while crewmates looked on and yelled their encouragement.

The Shuttle stored urine in storage in tanks that were periodically vented through space through a heated port in the hull. But On STS-41-D, a malfunctioning heater caused the material to freeze, creating a large yellow urinecicle outside the spacecraft. Astronaut Mike Mullane describes the aftermath:

[Mission control] called us. The heater on the outside of the dump nozzle failed, so this blob of frozen urine froze on the side of the vehicle, and what they were worried about was on reentry this blob of ice would break off, fly back, hit the tail, gouge out the heat tile. The tail would be burned off, and the Shuttle would crash: I thought of my life being threatened in many ways, but never by a block of frozen urine. Any rate, so we ended up using the robot arm. Hank used the robot arm to knock this piece of frozen urine off, but we were prohibited from using the urinal from that point, because they didn’t have any way of dumping it after that.

The crew spent the remainder of the mission peeing into sock-filled plastic bags, which they documented for posterity, along with the urinecicle itself:

The Shuttle toilet worked well on short missions, but sometimes ran into trouble with a full complement of crew. On one such flight, the overwhelmed toilet reversed flow and filled the cabin with a cloud of tiny freeze-dried fecal particles. It took several redesigns over the course of the program to finally solve the capacity problem, at which point it was time to pass the torch to the next generation of toilets on the International Space Station.

Since supplies flown to the station from Earth cost more than their weight in gold, there was early pressure on the ISS program to recycle water. The first Urine Processing Assembly, flown up in 2008, clogged up because dissolving bones gave astronaut urine a higher calcium content than the design anticipated. A modified UPA was later joined by a Brine Processing Assembly that could further desiccate processed urine. Together with a dehumidifier, these devices now recover about 98% of the water on the American half of the space station.

But for all that the urine system has advanced, fecal collection remains stuck in the Skylab era. Astronauts do their business in a one-time bag, seal their deposits (together with wipes, tissues, and gloves) inside, and then place everything in a hard-walled cylinder that eventually gets stuffed into a departing Dragon or Soyuz.

NASA has not been squeamish about gathering data to analyze every aspect of the space toilet. Used cylinders have been flown down from the space station and forensically disassembled, cataloguing each glove and wet wipe. NASA’s attitude to astronaut reports is best described as ‘trust but verify’.

Over the years, they’ve learned some things. NASA reports that a typical crewmember urinates six times a day, a number that seemed impossible until I started writing this essay and became hyperaware of bathroom visits. And a burst urine bag in the Skylab ground simulator taught the agency that ‘elimination events’ can vary in volume by a factor or three or more from the average. Peeing is not Gaussian.

The agency has also found that healthy people have pronounced variation in defecation—some need to go twice a week, while others go multiple times per day—and that cleaning up in space uses many more wipes and tissues than the same process on Earth, and requires a mirror.

All of this effort and toil has led to a space station toilet that, if not as beloved as a Japanese shower toilet, is at least tolerated by the crew.

Mars

Going to Mars takes space toilet challenges to the next level. The biggest of these is reliability: a space toilet is a complex device prone to breakdowns, and going to Mars would be the first NASA mission where a broken toilet could kill the crew.

Mars will need two very different toilet systems, one for the surface stay and one for the trip over. So it probably makes sense to talk about the two separately.

You may remember that the likely shape for a Mars mission is a six month journey out, followed by a 700 day surface stay, then a six month return trip to Earth. For radiation reasons, the crew probably has to spend all their time at Mars on the surface—you can’t leave one or two astronauts in orbit to tend the spacecraft.

The transit portions of this mission will resemble a stay on the International Space Station, which puts us within our experience base. But the 700 day gap, when the spacecraft has to wait in orbit with no humans on board, is a problem.

Both the water and sanitation systems on the space station are designed to be used frequently, and the regular flow of water through their parts helps to control microbial growth and flush away biofilms.

Once water sits stagnant inside a system for long periods of time, the fight against microbes gets challenging. NASA learned this lesson the hard way on the space station, when a drinking water assembly that waited too long on the ground arrived so contaminated with bacteria that it had to be thrown away. That system had been filled with sterile water, so you can imagine how much worse things get if you’re trying to stabilize a fecal tube or urine tank.

The Mars literature calls the problem of leaving habitable spacecraft unattended ‘quiescence’. You can think of it as the space version of getting a camper van or vacation home prepped for winter. There are multiple strategies that might work (filling the water and toilet systems with biocide, cooling the spacecraft to below freezing, draining and drying the system completely). But none has ever been tried, and it’s going to take some serious testing before we feel comfortable risking the life of a crew.

I’ve complained before that the runup to Mars would require long, boring human experiments in space. Designing for quiescence takes this problem to the next level. We need to build a space station, leave it empty for two years, then demonstrate that the toilet is not filled with cosmic horrors, and that all the life support systems can function for the six months it takes the crew to get back to Earth.

The other half of the Mars toilet puzzle is the surface stay. At 0.38g, Mars has the maximally inconvenient amount of gravity. On a log scale1, it sits right between zero-G and what we experience on Earth, with uncertain implications for toilet design. One thing we’ve learned from five decades of building space commodes is that there’s no substitute for actual flight experience. So it’s hard to see how we avoid building a rotating space station to test Martian plumbing in artificial gravity, along with other critical partial-gravity systems like dust filtration and smoke detectors.

Another design challenge for the surface is storage. If you include the mass of containers, four astronauts on a 700 day mission will leave behind three to four tons of highly septic waste. At least at first, this waste will give off gases that have to be vented to the atmosphere without allowing microbes to escape. And while this waste is bulky, you can’t just chuck it outside the airlock without making provisions to prevent it from contaminating the surface.

NASA has set itself the design goal of keeping astronaut waste sequestered for fifty years, and is in the early stages of testing vents and filters that can equalize pressure without getting rapidly clogged by dust. But this goal seems a little wild to me. NASA has trouble building structures that can last 50 years on Earth, let alone getting a level-4 biohazard storage shed on Mars right on the first try.

So a more workable approach might be to sterilize waste from the get-go. One promising approach is called torrefaction. It’s a kind of low-temperature roasting process that heats fecal bags to 200-250°C, hot enough to convert them to odorless char and recover some water while avoiding the problems (tar and noxious gases) that come with higher-temperature incineration. One potential design would combine fecal roasting with a trash compactor to create inert plastic tiles of processed waste that could be used as radiation shielding in the walls of the habitat.

Such are the glories of going to Mars!

Further Reading

• If you like quantifiable data, you are going to love the Body Waste Management NASA STD-3001 Technical Brief (2022)

• An enjoyable recent space toilet history from the Russian perspective: Overview of the Space Toilet History (2025)

• A very readable NASA technical brief on torrefaction: Torrefaction Processing for Human Solid Waste Management (2016)

• This paper on the toilet used in Artemis and the ISS goes into some of the details of what makes engineering the device so challening: NASA Exploration Toilet Hardware Technical Challenges and Accomplishments (2025)

Artemis II has just emerged from behind the moon, full of lunar cheer, so today I’d like to answer any questions people have about this mission in particular, the Artemis program it’s part of, or NASA’s Moon-to-Mars plans more broadly. A lot has changed at NASA in the past few months, but I have done my best to keep up and will do my best to answer!

No question too basic or silly; we’re all friends here.

Life arose very quickly on Earth. The biggest obstacle to nailing down a date is finding sufficiently old rocks to look for fossils in. Multiple lines of evidence point to life diverging from a little cell called LUCA (the Last Universal Common Ancestor) around 4.2 billion years ago, or soon after the oceans formed.

One thing we know about Luca is that she was a sophisticated organism, a strong, independent microbe containing the full machinery for assembling proteins from nucleic acid code, the shared inheritance of all life on Earth. Luca had a flexible metabolism, a large suite of proteins (and the genes encoding them), and even little luxuries like a rudimentary immune system.

In other words, Luca is not the kind of creature you expect to find wiggling out of the primordial ooze by chance. Scientists have tried to fill the gap between her and non-existence by positing a sequence of simpler biochemistries (called things like RNA world and RNP world) that bridge the gap between self-catalyzing molecules simple enough to arise spontaneously and the rich nanomachinery of modern life. Our current best guess is that cellular life evolved stepwise, with later stages eating or otherwise dismantling the evidence of their predecessors.

The problem with this theory is timing. The calendar is constrained on one side by the impact that created the moon 4.4 billion years ago, which surely sterilized the nascent Earth, and on the other side by the evidence of LUCA by 4.2 billion years ago. That timeline gives life less than 200 million years to get from nothing to the full suite of genetic code, transcription, translation, and incredible nanomachinery like the ribosome.

The timeline is especially fishy when you consider what happened after LUCA. The same chance processes that whipped up a ribosome out of nothing now dragged their feet for two billion years before inventing the eukaryote, a kind of luxury upgrade to the bacterium. And then it took another half billion years for some of those cells to start to aggregate, just a little bit, just to see what it feels like, finally giving rise to plants, animals, and all of our other modern problems.

There’s nothing wrong with evolution taking its time. But the contrast is odd. Why is waiting for complex life to emerge from simple life like waiting for a third My Bloody Valentine album, when simple life popped into existence so easily out of nothing?

Panspermia

Panspermia solves the riddle of abiogenesis in the best Douglas Adams fashion, by making it somebody else’s problem.

By the panspermia hypothesis, life showing up quickly on Earth should be no more of a surprise than mold showing up on a sandwich. Life arrived here fully formed from space, and got a foothold on Earth the moment the environment was hospitable enough to support it.

Our own natural history, of course, is full of these stories. Some jerk with a ship brought mosquitoes to Hawaii in 1826. The entire Australian continent is an object lesson in the danger of introducing foreign species. And the fact that we’re still arguing about who gave syphilis to whom in the 15^th century shows that it’s not always easy to tell the locals from newcomers.

The panspermia hypothesis is at least as old as Ancient Greece, though it may have arrived there from somewhere else. It comes in several different strengths:

• Mild: life arose somewhere else in the solar system, most likely on Mars or Venus, and arrived on a cooling Еarth via meteorite. In this family of scenarios, it’s possible for life to ping-pong back and forth between planets.

• Medium: life arose around some other star, and microbes hitch rides between solar systems on rocks and comets. In this variation, life would show up on habitable planets in the galaxy the same way it does on newly formed volcanic islands in the Pacific Ocean, by floating in naturally on the space breezes.

• Spicy: some alien Johnny Appleseed spreads life to new planets intentionally, the way Polynesian sailors used to leave pigs and yams on islands to make sure they’d have something to eat on the next visit. Maybe alien grad students are still broadcasting E. Coli out of passing UFOs; or maybe the whole process was set in motion by an extinct civilization, whose mindless machines carry on the inscrutable task of universal infection.

• Extra Spicy: in this version, life is omnipresent and spreads not just between solar systems, but between galaxies. For maximum heat level, you can combine it with Avi Loeb’s conjecture of an early warm universe scenario, where life arose in the warm cinders of the Big Bang before most stars and planets even had a chance to form. This version really puts the “pan” in panspermia.

A lot of people don’t like panspermia because it’s a bit of a cop-out. But not liking a hypothesis doesn’t refute it. And it’s not our fault if life arose somewhere other than our planet, and then made its way here. It’s a fact about the would that we need to establish, and the answer might help us understand the actual circumstances of abiogenesis.

A nice thing about the various panspermia theories is that they’re full of falsifiable detail. If you believe life spreads via comets, then it’s on you to catch a comet and show that it’s got microbes inside. If you think life had a Martian origin, then you had better go to Mars and start digging. This is what I ultimately like about it. Unlike ruminating about life origins, which involves building conjectural worlds in your mind, panspermia is full of experimental possibilities and plans for action. It gets you out of the house!

Let me go briefly through the different kinds of panspermia theory, starting from sober and ending with kooky.

Level 1: Lithopanspermia

Lithopanspermia is the most serious, buttoned-up version of the panspermia theory we have.

Imagine you’re a microorganism on ancient Mars. One day you’re eating sulfur inside a rock when there is a terrific commotion, the ground ripples, and you feel yourself floating into the sky. Soon it gets cold and you form a spore to wait things out. Hundreds of thousands of years pass. Then there’s another commotion, shaking, heat, and your rock cracks open on an unfamiliar world. It’s warm, it’s damp, and you find a piece of sulfur to eat under a rock. Life goes on.

The crux of the lithopanspermia idea is that microbes can live deep inside rocks, that it’s possible for those rocks to be ejected into space without undergoing sterilizing heat, and that their occupants can survive the long interplanetary journey and the heat of re-entry.

Somewhat remarkably, we have evidence for all of these being true.

The top five kilometers of our planet’s crust, including environments that have not had contact with the surface for millions of years, are inhabited by happy microbial communities. This underground biome, which we have barely begun to study, might account for something like 90% of life on Earth. And a lot of it is metabolically chill, capable of staying dormant for millions or (more controversially) hundreds of millions of years.

We know from computer models and meteorite finds that planets in the inner solar system have been swapping rocks for eons, via interplanetary splashback from meteor impacts. And magnetic evidence shows that the interior of at least one Martian meteorite found on Earth made the entire journey without ever being heated above 40°C.

Whether life actually crossed over on a meteorite is unknown, but we have a rough idea of how it could happen, and it’s something we need to test by going to Mars and other planets and moons to look for biosignatures there.

Level II: Life moves between stars

One problem with lithopanspermia as an explanation for life on Earth is that it doesn’t buy a lot of time. Mars is smaller than Earth, so it cooled faster, but that only gives LUCA a couple hundred million more years to evolve from nothing. If you want a more leisurely timeline for life origins, you have to think bigger.

Read more

Later that day, from the comfort of your hospital bed, you might come up with the concept of an ‘impact equivalent dose’ to quantify and relate the effects of the three blows you received. Maybe getting hit with the stick gave you an absorbed impact of 100 microcegłowskis, the golf club was 300 µCg, and the baseball bat was 550 µCg.

Once you’ve defined such a scale, you can use it to calculate the relative harm of other beatings by tallying up the various kinds of blows and where they land on your body. You might even derive some occupational exposure limits—perhaps 1200 µCg is the single-dose limit for a TikTok influencer, while office workers can only tolerate chronic daily beatings of 200 µCg, and small children shouldn’t be beaten at all.

This is the rationale behind units of effective radiation dose like the Sievert. They calculate the biological damage done by different kinds of ionizing radiation (gamma rays, protons, electrons, heavy ions) by applying a scaling factor to the energy that different parts of your body absorb1. The effective radiation dose will depend on what kind of ray or particle zapped you, how much of it hit you, and which specific tissues it passed through.2 And that final number will give you an idea of how likely you are to get radiation-induced cancer.

Effective dose models are calibrated against radiation exposures humans have encountered on Earth in the past (atomic bomb survivors are a favorite yardstick), and give a probabilistic measure of future bad outcomes like cancer and cardiovascular disease. The model in widest use today, called the Linear No-Threshold model (LNT), assumes that the relationship between dose and risk is linear, that the fundamental mechanisms of damage at the cellular level are the same no matter the type of radiation, and that there is no threshold value below which radiation exposure becomes harmless.

But to continue with my metaphor, imagine that one day instead of hitting you with a stick, I poke you with its pointy end. Now your effective dose model becomes less predictive. Because the mechanism of harm from stabbing is so different from blunt impact, the ‘effective impact dose’ you calculate may not capture the severity of the damage. You might even find that an otherwise light impact is rapidly fatal if it happens in the wrong place (like the jugular).

This is the situation scientists find themselves in with heavy ion radiation. Models of radiation equivalent dose assume that ionizing radiation either damages DNA directly, by barreling through a cell and breaking the strands, or through near misses, by creating reactive chemicals in close proximity to DNA strands. While this certainly happens with heavy ions (we observe some spectacular chromosome damage on astronauts returning from ISS), it leaves out an additional mechanism of damage that we are just beginning to discover, one that affects neighboring and distant cells that the ion didn’t pass through at all.

Non-Targeted Effects

“Non-targeted effects” is the catchall term for radiation-induced changes to cells that are not on the ionizing radiation track. They are a kind of ‘spooky action at a distance’ in the field of radiation biology.

One type of non-targeted effect is called the bystander effect, where cells in the local neighborhood of a cell hit by ionizing radiation will display levels of DNA damage similar to the cell that actually got hit. Just like you wouldn’t expect to find yourself in the hospital after seeing your neighbor got hit by a bus, the fact that this can happen is surprising. Bystander effects are also observed when irradiated cells are transplanted into new tissue (imagine losing your ability to walk because someone with a broken leg moved in next door to you), and when growth medium from irradiated cells is fed to a culture of pristine cells (imagine the same happening because you shared a burrito).

A second non-targeted effect is genomic instability, the tendency for the progeny and neighbors of an irradiated cell to show persistently higher mutation rates and chromosomal aberrations down the generations. This effect is different than the point mutations we expect to find in the classic model of radiation damage, and is also seen affecting cells far from the radiation track. It appears to involve changes in the epigenetic bookkeeping system that cells use to keep track of what genes they are supposed to be expressing.

Finally, there are abscopal effects that show up in remote parts of the body, hundreds or thousands of cell diameters away from the radiation track. Radiation oncologists since the fifties have observed cases where targeted radiation therapy in one body part will shrink or eliminate tumors far from the irradiated site, even though those tumors did not receive any radiation dose. Careful recent research has shown that abscopal effects are real, and are sensitive to factors (like what kind of ion is used in radiation therapy) that the standard model of radiation damage insists shouldn’t matter.

The conjectural mechanism behind all these effects is disruption to chemical signaling that affects whole cellular neighborhoods.

In some ways, the cells that make up our bodies are just like us—they are chatty, nosy, and susceptible to peer pressure. Something about heavy ion radiation disrupts the way they keep track of one another’s activities in a way that affects their development and gene expression. In the case of abscopal effects, it’s been shown that these disruptions are further mediated by the immune system, which can broadcast them to distant parts of the body.

Researchers are just beginning to map some of the complex chemical bookkeeping around gene expression that is responsible for these effects. As our understanding improves, the mechanism around non-targeted effects may grow clearer. But right now, we are stuck with a double riddle. We don’t know how epigenetic signaling and regulation work in healthy cells, and we don’t know how chronic exposure to heavy ion radiation in space would disrupt these mechanisms.

The uncertainty around the contribution of NTEs to radiation hazard is the main obstacle to calculating the radiation risk of a Mars mission.

Implications for Mars

There are five pieces of bad news about NTEs as they relate to exploring Mars.

Read more

I originally started compiling these statements in 2021 to maintain my sanity in the face of SpaceX enthusiasts debating about realistic timelines for Mars1, but figured the list might be of general interest to readers of this Substack. I’ve made it free to read for everyone.

I follow the timeline with some comments about Musk’s astonishing February 2026 pivot from settling Mars to building cities on the Moon.

2002

Musk’s public involvement with Mars begins with Mars Oasis, a piece of planetary performance art whose goal is to land a tiny greenhouse on Mars and photograph seeds as they germinate and begin to grow, a stunning image that Musk hopes will rekindle public interest in the space program.

Not having a rocket company at the time, Musk approaches the Russians for a pair of launch vehicles (repurposed ICBMs) to send this payload to Mars, but in the course of negotiations realizes that the real obstacle to space travel is prohibitive launch costs. He founds SpaceX to address the problem.

2005

At SpaceVision2005, Musk makes an early public mention of plans for the BFR2, a rocket design capable of launching 100 tons to low Earth orbit that will later evolve into Starship.

2008

SpaceX successfully launches Falcon 1, the first privately developed rocket to reach Earth orbit.

2009

In an interview at the Churchill Club, Musk says he will try to put an astronaut on Mars by 2020 or 2025.

2010

First orbital flight of SpaceX’s Falcon 9 rocket, which will become the most successful launch vehicle in history.

2011

Musk discusses a self-sustaining Mars base in an interview with Wall Street Journal’s Alan Murray. He anticipates putting a human on Mars in ten years [2021] in the best case, or 15-20 years [2026-31] in the worst case.

October: SpaceX and NASA publish a feasibility study on a concept called Red Dragon. In this mission design, a modified version of the Dragon capsule would launch on a Falcon Heavy booster and perform a retropropulsive soft landing on Mars before drilling underground for water.

2012

October: Musk makes the first mention of a large new SpaceX design called the MCT (later revealed to stand for Mars Colonial Transporter), a rocket previously known as the BFR.

Asked about a Mars landing in an interview on Nightline, Musk says “I'm confident at this point that it can be done. I think we'll be able to send, probably, the first people to Mars in roughly 12 to 15 years [2024-2027]. That's my estimate.”

2013

In an interview with Khan Academy, Musk says his best guess for a round-trip human mission to Mars is 2025.

2014

At the 2014 IEEE Aerospace conference, NASA presents an updated version of Red Dragon, this time envisioned as a sample return mission to launch in 2022.

2015

Musk tells blogger Tim Urban that SpaceX plans to “send an automated spaceship to Mars just to make sure you can send something there and back” by 2020.

December: First successful return and landing of a Falcon 9 booster stage after launch. This is an important proof of concept for reusable rockets, and a spectacular technical achievement by SpaceX.

2016

Starting in 2016, Musk and SpaceX begin making regular promises to launch an unmanned vehicle on Mars in the next synodic period, to be followed by human passengers two or four years later.

April: In a Facebook post and a tweet, SpaceX announces it plans to launch its first mission to Mars as soon as May 2018. After that “Red Dragon” flight, Musk says SpaceX’s goal will be to send at least one spacecraft to Mars during every interplanetary launch opportunity, which come every 26 months.

June: In an interview at the Code Conference, Musk says “we should be able to launch people [to Mars] probably in 2024 with arrival in 2025. […] That’s the game plan, in 2024, to launch the first of the Mars Colonial Transport system with people. This will be a very big rocket.”

September: In an interview with Sam Altman (then at YCombinator) Musk says a Mars landing “is potentially something that can be accomplished in about ten years [2026], maybe sooner. Maybe 9 years [2025].” He cites keeping SpaceX solvent as a key challenge to his Mars plans.

September: In a major presentation at the International Astronautical Congress in Guadalajara, Musk rolls out the “Interplanetary Transport System”. Musk makes it clear that his goal will be to launch at every Mars transfer opportunity, beginning with Red Dragon in 2018, and transitioning to the Mars Colonial Transporter whenever that becomes operational. He estimates that it will take 20-50 Mars launch windows (50-110 years) from the initial landing to get to a Martian population of 1,000,000.

2017

February: SpaceX announces that Red Dragon will slip from the planned 2018 launch date to 2020.

SpaceX signs up Yusaku Maezawa for a trip around the Moon in a Dragon space capsule, to be launched in 2018 on a Falcon Heavy booster.

June: In an interview with the Washington Post, Musk stresses his intent to launch rockets to Mars at every opportunity beginning in 2020. “Essentially what we’re saying is we’re establishing a cargo route to Mars… It’s a regular cargo route. You can count on it. It’s going happen every 26 months. Like a train leaving the station. And if scientists around the world know that they can count on that, and it’s going to be inexpensive, relatively speaking compared to anything in the past, then they will plan accordingly and come up with a lot of great experiments.”

July: Musk publicly abandons the Red Dragon concept in favor of the larger rocket laid out in his IAC talk the previous year.

September: Speaking at the IAC congress in Adelaide, Musk presents an updated vision of his 2016 plan. An unmanned landing by cargo ships will depart for Mars in 2022, followed by the first human astronauts in 2024.

2018

Musk publishes a paper in New Space titled “Making life interplanetary”. The paper, an edited transcript of his IAC talk, lays out an architecture for reaching Mars with a large reusable rocket (again called the BFR) capable of launching 150 tons to low Earth orbit. The cornerstones of the architecture are reusability, frequent launch cadence, orbital refueling, and fuel factories on Mars.

Musk predicts the first cargo landing on Mars will take place in the 2022 launch window, followed by two cargo and two crew landings in the 2024 synodic period.

At this stage, the shape of the return journey is nebulous. The first arrivals on Mars will also become settlers, working on a pilot plant for creating rocket fuel from local resources that would enable their eventual return to Earth (if they choose to go back).

Musk’s vision is of an initial landing site accruing Starships and crew members with each launch window, eventually expanding to become the nucleus of the first Martian city.

February: Musk puts his personal Tesla roadster on the maiden flight of the Falcon Heavy booster, sending it out towards the orbit of Mars.

September: plans for the dearMoon circumlunar mission with Maezawa are updated. Instead of launching aboard a Dragon capsule on Falcon Heavy in 2018, Maezawa and a group of guests will fly in 2023 on a human-rated version of Starship. Maezawa is reported to be fronting a ‘meaningful’ portion of the Starship development cost.

2019

September: Musk says Starship, which has yet to fly, “could be sending people to orbit within a year [2020].”

2020

May: The launch of Crew Dragon Demo-2 sends a pair of NASA astronauts to the International Space Station on a Falcon rocket. It is the first manned flight by SpaceX, the first manned flight on an American rocket since the Space Shuttle, and the first orbital flight by a commercial provider.

December: In an interview with Mathias Döpfner, CEO of Axel Springer, Musk says he's ‘highly confident' SpaceX will send humans to Mars in 2026, and possibly as early as 2024 “if we get lucky.” Musk adds: “And then we want to try to send an uncrewed vehicle there in two years [2022].”

2021

April: NASA enlists SpaceX to develop the Human Landing System, a derivative of its Starship second stage that will serve as a lunar lander for the agency’s Artemis project.

September: SpaceX launches future NASA Administrator Jared Isaacman and a crew of three on Inspiration4, the first private orbital spaceflight.

December: Asked on the Lex Fridman podcast when SpaceX will land a human being on Mars, Musk takes a long moment to think and replies, “best case is about five years [2026], worst case ten years [2031].”

2023

April: First flight test of Starship.

October: Speaking by videoconference at the IAC congress in Azerbaijan, Musk says he anticipates an unmanned test landing on Mars within the next four years [2027].

2024

June: Maezawa cancels the dearMoon circumlunar Starship flight, citing lack of a clear near-term schedule.

September: Musk tweets “The first Starships to Mars will launch in 2 years when the next Earth-Mars transfer window opens. These will be uncrewed to test the reliability of landing intact on Mars. If those landings go well, then the first crewed flights to Mars will be in 4 years. Flight rate will grow exponentially from there, with the goal of building a self-sustaining city in about 20 years. Being multiplanetary will vastly increase the probable lifespan of consciousness, as we will no longer have all our eggs, literally and metabolically, on one planet.”

2025

January: First flight test of Starship, version 2.

Responding to conjecture that the Moon might serve as a source of propellant for SpaceX’s Mars missions, Musk tweets “No, we’re going straight to Mars. The Moon is a distraction”.

March: Musk tweets “Starship departs for Mars at the end of next year, carrying [humanoid robot] Optimus. If those landings go well, then human landings may start as soon as 2029, although 2031 is more likely."

May: Musk gives an update on the progress of Starship at Starbase. The plan is now to send five Starships to attempt a Mars landing in the 2026 launch window. The first Starships will land an Optimus robot, with human landings to follow in either 2028/9, or more likely the 2031 window. Musk says he plans to land five Starships on Mars in 2026, 20 in 2028/9, 100 in 2031, and 500 in 2033.

December: Bloomberg reports that SpaceX is planning to go public in 2026, at a valuation of $1.75T.

2026

February: SpaceX acquires xAI.

Musk tweets “For those unaware, SpaceX has already shifted focus to building a self-growing city on the Moon, as we can potentially achieve that in less than 10 years, whereas Mars would take 20+ years.

The mission of SpaceX remains the same: extend consciousness and life as we know it to the stars.

It is only possible to travel to Mars when the planets align every 26 months (six month trip time), whereas we can launch to the Moon every 10 days (2 day trip time). This means we can iterate much faster to complete a Moon city than a Mars city.

That said, SpaceX will also strive to build a Mars city and begin doing so in about 5 to 7 years [2031-33], but the overriding priority is securing the future of civilization and the Moon is faster.”

Responding to pushback about his change of focus, Musk clarifies: “We’re still going to Mars and the timeframe for building a self-growing city there is still about the same at 20 to 30 years. [2056-2066]”

He later tweets: “To be clear, we are still going to do Mars. I don’t think this change affects the time to a Mars city being self-growing by more than 5 years and it might turn out to accelerate Mars.”

Remarks

Several patterns emerge in Musk’s Martian promises, besides the obvious one of a Mars landing always receding over the horizon.

One pattern is a reliance on spectacle to motivate and generate excitement. The Mars Oasis project is an early example, but others follow: stacking Starship for public display two years before it is ready to fly, the iconic photograph of a spacesuited astronaut passing Earth in a Tesla, or the lavish SpaceX renderings and movies that serve as a backdrop to Musk’s major public announcements about Mars.

Another pattern is Musk’s irrepressible desire to go big. The modest Red Dragon (which Musk could launch with his own money tomorrow, if he wanted to) gives way to Starship v1, which then gives way to Starship v2, v3, and now v4. By the time a previously promised rocket is ready to perform, Musk has typically lost interest and moved on to bigger plans.

A third, unexpected pattern is Musk’s steadfastness. Despite his aspirational mendacity in talking about timelines, until this year his vision for settlement remained rock steady. Mars was the backup planet we needed to colonize to preserve ‘the light of human consciousness’, and it was important to create a sufficient industrial base there for a colony to be self-supporting (and capable of developing interplanetary travel) even if they lost contact with Earth.

Just as consistent was his technical roadmap. Since at least the 2016 presentation to the IAC, Musk had stressed that SpaceX needed to develop four technologies (cheap launch, rapid reusability, orbital refueling, and fuel production from atmosphere and ice) to get to Mars. The company seemed well on its way to developing the first two capabilities, and was committed to inventing the third by its HLS contact with NASA.

Musk was also always clear that the purpose of SpaceX’s commercial arm was to subsidize his Mars plans. While cheap launches on Starship would revolutionize access to space, colonizing Mars was a grand civilizational project, not a money-making effort. For that reason, SpaceX could not become a public company before reaching Mars, because there would be no way to justify the Mars project to shareholders or a board of directors on financial grounds.

This unusual steadiness is what makes Musk’s pivot to the Moon last month so astonishing. By merging SpaceX with a company journalist Mike Isaac calls “a conveyor belt that carts cash into a furnace”, Musk has made it impossible for SpaceX to carry out his founding vision. Whatever the prospects for AI in the long term, it will have an infinite appetite for capital in the near term, and the commercial side of SpaceX cannot earn enough to fuel two money bonfires at the same time.

The structural changes Musk is making to his space company are permanent. Even if Musk later changes his mind about Mars, SpaceX is now welded to an enterprise that can immolate arbitrary quantities of money, as well as to an IPO that (by Musk’s own earlier logic) will preclude them from carrying out the founder’s great vision.

For true believers, this is a bitter moment. Whatever your private multiplier for Musk’s timelines, his willingness to spend vast money on Starship, and the astonishing technical progress of that program, served as a kind of proof of earnestness for his Mars plans. Even if you weren’t on board with the colony, the prospect of cheap Starship cargo flights to Mars created dizzying prospects for exploration.

Musk’s obsession also created a reality distortion field that made it difficult to discuss alternatives. As long as SpaceX was promising landings in two years’ time, it made little sense to advocate for a Mars mission by NASA (who can’t get even get back to the Moon), or to push for robots to be sent on rockets that would be made obsolete by Starship the first time it managed to refuel its tanks in orbit.

The rugpull must hit particularly hard for engineers at SpaceX who devoted their professional lives to realizing a vision that turned out to be a chimera. Like their predecessors at Tesla, they’ve learned that Musk can be fickle, and even years of highly public commitments to a transcendent vision are no guarantee of his future interest.

Further Reading

1. Scott Manley has a nice video about the Mars Oasis project.

2. For a 2015 snapshot of Elon Musk’s thinking about Mars colonization, see this long but informative series by Tim Urban.

3. For an independent evaluation of Musk’s Mars architecture, see About feasibility of SpaceX’s human exploration Mars mission scenario with Starship [2024]

Okay!

But as ever in space flight, when someone says “just do X”, there’s been 40 years of applied research into X, and taverns up and down the Space Coast are filled with grizzled engineers who devoted their careers to X and are drinking to forget. So let’s talk about why no one in sixty years has ever built a spaceship you can tap-dance in.

Research in artificial gravity has followed a kind of bathtub curve. At the start of the Space Age, everyone assumed that humans would need gravity to do any serious work in space, and there was a lot of foundational work around the minimum requirements for a rotating space home. Then, in 1973, Skylab showed that crews could spend multiple months in free-fall, get things done, and return to Earth without apparent harm. And so interest in artificial gravity waned. For a while, it looked like humanity might get away with exploring the entire solar system with just our socks on.

But over the ensuing decades, we learned that prolonged stays in zero-g do have health effects, and that these are both pernicious and intractable. So artificial gravity is back, baby! And because orbital mechanics forces missions to Mars to last multiple years, they are the obvious candidate for trying it out.

Basic Principles

The idea behind artificial gravity is simple. If you spin a large structure in space fast enough, anyone standing on its edge will feel a force in the outward direction. If the structure is big, and rotates slowly, that force will be hard to distinguish from the kind of gravity we have at home, and the crew can go jogging on the coolest track in the Solar System.

How much force you feel on the rim of a rotating habitat depends on both the spin rate and radius. At one revolution per minute, a structure needs to be nearly two kilometers long to create 1g. At ten revolutions per minute, you only need a structure 18 meters long, about the length of a semi truck. So it is important to know how much spin we can get away with.

As with many aspects of human space flight, it all comes down to barfing.

Read more

Within a few days, an astronaut’s body starts to to resorb bone, particularly in the hips and lower spine, a process that will continue at a steady rate for as long as they stay in space. There are marked changes in the immune system, with latent viruses (like chickenpox and Epstein-Barr) are likely to reactivate. And at longer time scales, 70% of long-duration crews develop a poorly understood degenerative condition of the retina and optic nerve with no analogue on Earth.

The multiple health impacts of prolonged weightlessness are collectively called deconditioning. In many ways they mimic an accelerated aging process, and there are worse ways to think of a Mars mission than a rocket-propelled nursing home trying to get its crew back into gravity before their bodies give out.

In this post, I want to focus on the two biggest hypogravity-related health risks facing a Mars-bound crew: damage to the bones and eyes.

Bones

The problem of bone health is emblematic of what makes space travel a drag. It’s not a virile manly problem like making sure your rocket stays erect, but one of slow and inexorable physical decline, remarkably similar to aging.

Human beings in free fall can expect to lose about 1% of lower body bone density each month. When you measure bone strength instead of density, the decline is sharper, around 2.5% per month in the hip joint. Unlike other adaptations to space, the rate of bone loss does not level off with time. Extrapolating from bedrest analogues suggests astronauts on a thousand-day mission could lose over a third of their hip and femur bone mass. This much bone loss would put them at extreme risk of fracture, even in Martian gravity.

One study that examined the bones of 14 returnees from the space station found that “the post-flight proximal femoral strengths for three of the astronauts are close to the median strength for elderly Caucasian women, the population at greatest risk for osteoporotic fracture.” One of the group lost so much strength in his femur that a stumble or fall onto a carpeted floor would have broken his hip.

In other words, ISS astronauts have already been coming back to Earth with old lady bones. The rate of bone loss on a six-month ISS mission is comparable to the decline a man sees between age 40 and 60, and is about six times greater than what women experience during menopause.

Tracking bone loss

In 2010, NASA convened a Bone Summit to come up with a better plan for mitigating bone loss in their crews.

Historically, NASA measured bone loss using an imaging technique called DXA (dual‐energy X‐ray absorptiometry), which shines X-rays through bones to give a rough measure of their total mineral density. NASA likes DXA because it is cheap and ubiquitous, being the main clinical tool for measuring bone loss in the elderly.

But DXA is a measure of bone density, not strength, and astronauts are not representative of the elderly population that DXA scores are calibrated against. Since the specific risk we’re worried about is bone fractures, we need to a way to directly measure bone strength.

Large bones in the body get their strength from a foam-like core of mineralized tissue called trabecular bone. As the body resorbs this bone, the links in its latticework structure get thinner. If they thin to the point of breaking, the lattice will never re-form, and the bone becomes permanently weakened.

Tracking changes in trabecular bone means feeding a 3D scan of each astronaut’s bones into a computer model, a complicated method without a lot of history. NASA has just started conducting these scans systematically; the Bone Summit insists that they can’t offer a proper clinical opinion until there have been multiple long-duration flights under the new protocol.

But the same summit suggested a preventative drug that has so far shown remarkable results.

A magic bullet?

In the past, NASA’s chief weapon against bone loss has been exercise. Astronauts on the ISS spend hours a day working out on a kind of orbital Soloflex called the Advanced Resistive Exercise Device. Together with better nutrition, these workouts have brought the rate of bone loss to about half of what it was on Mir, but at the cost of about three hours a day in useful time crew time. A single ISS crew member spends more time exercising each week than the entire space station has available for scientific research.

Read more

This fact makes people mad, because there are all kinds of other obstacles that are fun to solve (orbital refueling, landing heavy payloads, making rocket fuel out of Martian air), and life support has a fun factor of zero. It is a thankless world of dodgy sensors, failing bearings, and bacteria trying to grow on absolutely everything.

But if we want to get to Mars alive, we need for this stuff to work.

Requirements

An astronaut in space needs 840 grams of oxygen, 2.8 kilos of water, and 1.8 kilos of dried food a day to stay alive. They also appreciate the little touches, like water to wash with (0.7 kg), fresh clothes (1.5 kg per week), wet wipes (0.2 kg/day) and a toilet (1.4 kg for canisters and wipes).

On the output side, each astronaut exhales around a kilo of carbon dioxide and pees out a liter and a half of urine. They also produce a fairly small quantity of feces and menses (though many women opt to medically induce amenorrhea during space flight).

On shorter missions like Apollo (~12 days) or Shuttle flights (~14 days) it makes sense to pack everything a crew needs with no attempts at recycling. In this paradigm, carbon dioxide can be scrubbed from the cabin air with disposable lithium hydroxide cartridges; everything else is carried along in the space version of a picnic basket.

On longer missions, trying to carry single-use supplies gets unwieldy. A crew of four on a 1,000 day mission to Mars would need 48 tons of consumables, about equal to the mass of the entire spacecraft1. And even if mass were unlimited, there simply wouldn’t be enough room to fit everything on board.

So past a threshold of about 30 days, you have to make some attempt at recycling.

Water

From the diagram above, you can see water is the obvious place to start. Four to five kilos of the stuff flow through each astronaut every day, and you can reclaim most of that amount just by condensing cabin air on a cold surface in the spacecraft.

On the Mir space station, this used to happen organically. Collecting water was a grubby job that involved chasing beach-ball sized spheres of condensate around the colder parts of the spacecraft with trash bags before they could climb into the walls and cause mayhem. Crew members spent three to four hours a day on this dirty and difficult task.

The ISS, a more civilized place, uses a heat exchanger that you can think of as a space-rated dehumidifier. It collects water and reprocesses it for drinking.

Once you’ve squeezed the cabin air dry, the next best source of reclaimed water is urine. Gravity makes it easier to boil astronaut pee on Earth than in space, but a heated centrifuge weighing nearly a ton gets the job done on ISS at about 87% efficiency.

(The limiting factor in urine distillation is actually the high level of calcium from disintegrating astronaut bone, a nice example of how problems in space find ways to compound one another.)

Once the pee is boiled, you’ve reached a 90% level of water recovery. Most of what remains is locked away in an unpleasant substance called urine brine. The ISS has been testing a contraption that dries this out using a system of permeable membranes, but the residue is not easy to handle. To quote a NASA study:

“When wastewater brines are dried, the residual is inevitably a viscous goo, laden with particles of precipitated solids. This brine residual causes several problems for traditional recovery systems, such as clogging pitot tubes, causing bearings to seize, and fouling heat transfer surfaces.”

Drying this enjoyable substance gets you to ~97% closure, at which point you’re left chasing the 90 grams of water trapped each day in astronaut feces.

For NASA, this remains the final frontier. Solutions for unlocking this treasure have ranged from slow roasting, to charring, to simply putting the fecal canisters in a microwave (though presumably not the one in the ship’s break room).

Recovering fecal water puts you near 98% closure, with most of what’s left lost through trash disposal (wet wipes and food packaging).

The water system illustrates the problem of diminishing returns in regenerative life support. Getting near the 98% recovery rate NASA targets for Mars takes four pieces of heavy machinery (the heat exchanger, urine still, brine dryer, and fecal oven) that recover 64%, 26%, 7%, and 1% of cabin water respectively.

Even though half of this machinery recovers only 8% of the water, the crew will be just as dead if any of it stops working.

This trade-off between closure and mechanical complexity comes up repeatedly in life support design, with the added caveat that all the subsystems interact with each other and the bodies of the crew, so they can’t be tested in isolation.

Air

The air system on a spacecraft has a number of jobs to do. It has to replenish oxygen, remove carbon dioxide, and filter out airborne contaminants ranging in size from viruses to hammers. It also has to keep the air in the spacecraft uniformly mixed, detect a variety of toxic gases, and let someone know if the spacecraft is on fire.

Carbon dioxide is what kills you first if the air system fails. A CO₂ concentration much above 5% (50,000 ppm) will incapacitate a crew, meaning that whatever removes the gas from cabin air has to be pretty reliable and not break for too long. (Astronauts have given themselves bad headaches just by falling asleep in a corner of the spacecraft that had inadequate ventilation).

The early space program through the Shuttle era scrubbed carbon dioxide with disposable canisters of lithium hydroxide, a powder that chemically binds the gas. On the ISS, CO₂ is removed by a machine about the size of a car engine that pumps dried cabin air over a bed of reusable zeolite beads. These beads adsorb CO₂ on their surface but do not react with it chemically. Once a bed is full, it is shunted to vacuum and heated until the trapped gas vents off into space.2

Scrubbing CO₂ out of cabin air poses a special challenge because there is no distinct waste stream to process. The crew exhales into the same ambient air they breathe in, and it’s thermodynamically hard to get all the gas out of a dilute solution. The scrubber on the space station struggles to bring CO₂ levels below 3500 ppm, about nine times the baseline level on Earth, and three times more than you might encounter in a stuffy office. Whether this affects astronaut performance depends on whom you ask (the Americans say definitely; the Russians, whose paycheck gets docked for whining, say it’s no problem).

In comparison to CO₂, replenishing oxygen is straightforward. You can make all the oxygen you want by running an electric current through water. But the equipment for doing this is heavy, and leaves gaseous hydrogen as a byproduct. The newly separated H₂ and O₂ want desperately to reunite back into a molecule of water, and thwarting this desire takes a fair amount of mass and engineering. So in many mission designs, it can make sense to just forget recycling and treat oxygen as a consumable.

Food

Food may be my favorite technical barrier to Mars travel, because everyone assumes it has been solved, or that it is easy to solve, while the people working on it mop the sweat from their brow during the day and try keep the shaking in their hands from rattling the ice cubes in their whisky glass at night.

Astronauts have hated space food ever since the first meat cubes came back uneaten from Project Gemini. Even on the ISS, where fresh foods are often available, getting crews to eat adequately is a struggle. Whether it’s because the stomach senses satiety differently in zero gravity, or because the space station smells like a toilet, crews have historically consumed only 80% of their rations.

On a multiyear mission, such a calorie level would lead to malnutrition and embarrassing deficiency diseases like space scurvy. So we need to come up with ready-to-eat meals that are nutritious, storable for five years without refrigeration, and appetizing enough that a crew can eat them for a thousand days without wanting to murder each other.

These kinds of meals don’t exist. Their closest equivalent is the military meal ready-to-eat (MRE). But as any soldier or prepper will tell you, an MRE is not something you can subsist on. The meals are not nutritionally complete, and soldiers’ own backronym for the combat ration (meals refusing to exit) sheds light on a notorious shortcomings. Defense department guidelines stress that soldiers should not be fed MREs for more than 21 days at a time.

The absence of cold storage on a spaceship might surprise you. Surely the one thing space doesn’t lack is coldness. Why not just hang six tons of frozen burritos out the airlock and be done with it?

But a spacecraft is more like a thermos than a freezer. Everything in the main pressure vessel will tend to come the same temperature, so a special cold storage area would need to be outside crew capsule. Not only does this mean regularly using the airlock (an incredibly risky process that wastes air) but it would deprive the crew of the considerable radiation protection afforded by sleeping behind a wall of chicken tetrazzini.

So a space root cellar is out. What we’re left with on the spacecraft are the cooking facilities of an office break room: a microwave, hot water tap, and a mini fridge.

Growing your own

Of course on Earth, this problem has been solved for ten thousand years or more . Plants make tasty food on demand, they recycle a number of astronaut waste products, and they can really spruce up a spacecraft. So why not take some with us?

But for all the appeal of subsistence farming as a lifestyle, it comes with problems. We know next to nothing about zero-g agronomy, and we’ve already lost some space crops to mold. Making the crew dependent on plants puts astronauts one bad harvest away from starving to death.

Plants also need a huge amount of room to grow, and their inedible parts have to be composted. Doing this has proven challenging even in closed-cycle farming experiments on Earth, so NASA quite rightly doesn’t want to add it to all their other problems in space.

So while a Mars crew will likely try growing plants to supplement their menu, they can’t rely on them for core nutrition. One way or another, we need to find a way to store room-temperature food for up to five years without it turning into an unappetizing brown mush.

Body Waste

The space toilet has been a reliable source of fascination for every rising generation of space nerds. Years of engineering effort have reduced the humiliation level of the device (Shuttle astronauts had to practice centering their anus in the cross-hairs of a video camera) but releasing the hounds on the ISS remains a much less joyful experience than, say, using a Japanese shower toilet for the first time.

On the space station, feces gets vacuumed into a bag full of antimicrobial powder. This bag is then sealed and placed in a cylindrical canister for long term storage, until it can be sent down to Earth.

Only the most daring life support engineers want to crack these cylinders open and try processing the stuff. But storing it on a Mars mission is a challenge because the containment cylinders take up a lot of space. Long-term storage is a special worry on the Martian surface, where the bacteria in human waste pose a serious potential contamination hazard, and have to be prevented from leaking into the environment over timescales of many decades.

Medicine

Early space medicine was a zero-gravity version of “walk it off”. Apollo doctors filled the spacecraft’s glove box with amphetamines and tranquilizers and wished the crew a good flight.

As missions grew longer, the medical kit expanded. Mir had a medical case of surgical instruments with a bilingual instruction manual, and often flew a doctor as one of the crew. The ISS is practically an orbiting Walgreens, offering residents the use of an ultrasound machine, defibrilator, pregnancy test, and a full suite of pharmaceuticals from aspirin to Zoloft.

Being long and remote, a Mars mission will require a medical bay that can handle health risks inherent to the mission (bone fractures, vision loss, toxic burns, cancer) as well as unexpected illness. While a spacecraft might seem like the last place to worry about infectious disease, the immune system gets weird in space and old viruses can reactivate. At the same time, the crew is sharing their living space with a YOLO experiment in fungal and microbial evolution in a highly mutagenic environment.

Even though ISS astronauts pop pills like candy, our knowledge of how pharmaceuticals behave in space is minimal (there have been just two experiments). We’re also in the dark about how long a shelf life medication has in space, which is a particular issue for things like antibiotics.

When it comes to administering anesthesia or doing surgery, knowledge is purely theoretical, and limited to common sense advise like not using anesthetic gases in a confined space.

A problem unique to spaceflight is figuring out the downstream effects of drugs and their metabolites on the life support system, which is full of sensitive components like catalyst beds that can be poisoned by small quantities of material. There’s little point in figuring out how to administer chemotherapy in space if it’s going to brick the water purifier and kill the whole crew.

Fire

Fire doesn’t usually get lumped in with life support, but since not being on fire is essential to the crew’s survival, I like to throw it in.

The worst space fire to date happened in 2003, when a defective oxygen canister on Mir ignited and burned for fourteen minutes, spraying droplets of molten metal clear across the spacecraft. The crew emptied three fire extinguishers and sustained second-degree burns in a futile attempt to put it out, but ultimately they just had to wait for it to finish burning on its own.

Though NASA and Roscosmos downplayed the incident, this was a very close call. Had the fire started a few days earlier, before a departing Shuttle had emptied the module of trash bags, it would likely have spread throughout the station. As it was, the limited fire filled Mir with smoke and nearly forced the crew to evacuate.

Fire scares on the ISS have so far been more benign, but any long-duration spacecraft is by its nature a fire hazard. A Mars-bound spacecraft will be a crowded environment with miles of wiring and significant heat sources (the zeolite beds in the carbon dioxide scrubber need to be heated to 400°C, for example). And because there’s no way to cut a mission short, the crew will not just need to be able to extinguish a fire, but live in its ashes for up to two years.

This means that the life support system has to be able to absorb the combustion products of a major fire and keep working, or at least be fixable afterwards. That’s a heavy design constraint to adhere to, especially when you can’t simulate an actual microgravity fire on Earth.

Exercise

Exercise is an odd duck. NASA requires space station crews to work out for multiple hours a day to combat both bone loss and muscle atrophy. This seriously eats into available crew time while stressing the life support system.

There is a class of drugs called biphosphonates that inhibits bone loss, but NASA screwed up the relevant experiment and forgot to set up a control group who would take these drug in space without exercising. So we still don’t know if drugs alone provide adequate protection against bone loss in space, and ISS crews still spend hours each day working out like gerbils on a treadmill.

Culturally, it doesn’t help that the astronaut corps is a bunch of health nuts who just love to work out. There is a pronounced tendency in the literature to assume that exercise is always good, and horror at the idea of turning our astronauts into a bunch of zero-gravity layabouts.

But it’s very important to find out whether we can get away without exercising during a Mars surface stay, since the amount of mass we can land on the planet is a serious constraint on habitat size. Having to include exercise equipment means limiting available exploration time, packing more food and oxygen, and upsizing the life support system.

Laundry

Laundry on the ISS is handled with disposable clothing, which the astronauts try to stretch for as long as possible. Astronaut Daniel Petit explains the drill:

We were stretching [changing clothes] out to about every 8 to 10 days, and the indicator that it was time to change your underwear would be when you started to get a rash around your waist, and then it was time to change that pair of underwear. Then you would down-mode it. You’d wear your underwear for prime daytime underwear, then you’d down-mode it to exercise underwear. We’d exercise for two, two and a half hours a day, and you’d use a different set of clothes for exercising. So, your day clothes would be down-moded to exercise clothes, and then from there you’d down-mode them to rags for cleaning up messes.

Sending disposable clothing to Mars means figuring out where to put it all without making astronauts feel like they live inside an old closet. New clothes also gunk up the air filters with lint.

Washing clothes, on the other hand, creates a new wastewater stream for the life support system to handle. When you remember that substances like hair conditioner and lotion have historically crippled the water system on the ISS, you can understand the apprehension that a couple of Tide pods inspire in the engineering department.

The laundry problem is of course exacerbated by exercise. And on a Mars surface mission, it may be further exacerbated by the need to handle dust.

NASA’s strategy for the time being is to search for fabrics that don’t soil easily and try to develop a kind of superFebreeze for space. But like the food problem, the laundry issue is both unglamorous and a fairly serious capability gap.

Stowage

The final problem I’ll mention in connection with life support is stowage—where to put things. We tend to look at life support systems in terms of mass balance, but in many cases the more severe constraint on a spacecraft is volume.

While the ISS is enormous by space standards (it has about the volume of an airliner), its inside is still a warren of cabling and equipment, much of which folds into the walls and requires hours of prep work to get access to.

Cargo bags sometimes leave entire areas of the space station off limits, and planning and managing how to Tetris all this stuff together and move it around is a full-time job for logistics personnel on the ground.

Just keeping track of what is on the space station, and where it’s hiding, is a challenge. Though items are tracked almost obsessively, down to telling the crew which trash bag to put a discarded item in, mission control routinely sends up WANTED posters of missing items. There’s an informal contest among the astronaut corps to see how long an object can stay lost on the space station (the record so far is eight years).

Because of size constraints, stowage on a Mars spacecraft will be much harder. The main pressure vessel has to fit inside a rocket fairing, putting strict limits on available volume.

Inside that space, the crew will need to have ready access to all equipment that might break, whatever spare parts they need to fix it, and still find room for three years’ worth of supplies, sleeping quarters, space suits, a toilet, and a gym.

Further Reading

1. For more on wrangling water globules aboard Mir, I highly recommend David Wolf’s entertaining oral history

2. To read about the challenges of removing carbon dioxide in ISS, here’s a 2024 status report on the latest tech being tested there.

3. For a deep dive on complexity vs. recovery tradeoffs in water reclamation, see Water Recovery Trades for Long-Duration Space Missions

4. On the question of space food, see Progresses [sic] in processing technologies for special foods with ultra-long shelf life DOI: 10.1080/10408398.2020.1853034

5. A 2025 paper on the latest efforts in creating a space laundromat.

6. For a real deep dive into the issues I raise in this post, see Logistics Needs for Future Human Exploration Beyond Low Earth Orbit (2017). DOI 10.2514/6.2017-5122

The study follows up a 2025 result that detected long-chain organic molecules in a sedimentary rock formation at Gale Crater. That study found three alkanes (decane, undecane, and dodecane) present in a rock sample at the 30-50 part per billion level.

The rock layers these compounds were found in formed ~3.5 billion years ago1 and remained buried deep underground until 80 million years ago, when wind erosion exposed them at the surface. The new study makes a quantitative assessment of just how much of these compounds must have been originally present to leave the residual amounts observed by Curiosity.

The answer is “piles and piles of the stuff”, a result that is easy to explain if you have a lake teeming with microbes, but hard to reconcile with the known abiotic pathways for getting organic compounds into Martian mud.

Since the result is so significant, I thought I’d go through it in some detail.

Welcome to Gale Crater

Gale Crater was chosen as the landing site for the Curiosity rover in 2012 in the correct hope that it would turn out to be an ancient lake bed. Years of roving and drilling there have led to exciting discoveries. Thanks to Curiosity, we know that Gale Crater was home to Earth-like freshwater lakes for at least tens of thousands and potentially millions of years, and we know that that sediments there contain complex organic chemicals consistent with the presence of ancient life.

The Cumberland drilling site where alkanes were found in 2025 is a mudstone, which is exactly what it sounds like. Muck gets deposited on a lake bed, and over time the layers compact down and harden into rock. For three billion years or so, this rock stayed peacefully buried, until it was unmarsed by the wind in the recent geological past.

Since Mars has no magnetic field and a thin atmosphere, cosmic rays smash into any rock that is close to the surface. These impacts leave behind trace isotopes that act as a kind of clock. In the case of the Cumberland mudstone, three isotopes (³He, ³⁶Ar and ²¹Ne) give independent values for this exposure time that are in good agreement with one another. The mudstone has been exposed to cosmic rays (that is to say, near the surface) for about 80 million years (78 ± 30 My).

Except for that last interval, the rock has led a sleepy existence that has preserved the evidence of its formation remarkably well, given its enormous age.

What are alkanes?

Like the mudstone they’ve been resting in, the alkanes in the Cumberland sample are sleepy compounds—simple hydrocarbon chains 10-12 atoms long. The shortest of them, decane, is a common constituent of gasoline. Here’s what it looks like in a Kekulé diagram:

(If you’ve never seen one of these diagrams before, each vertex represents a carbon atom, and the lines are single bonds connecting them. The 22 hydrogen atoms needed to fill out the molecule are not shown.)

Alkanes are close relatives of molecules called fatty acids, which have a similar carbon skeleton with a carboxyl group at one end. For example, the fatty acid analogue of decane is decanoic acid, a constituent of coconut oil:

Unfortunately, because of how the samples are heated on Curiosity, the rover can’t tell the difference between alkanes and their fatty acid equivalents. And because of how long these compounds have been entombed, it’s possible that the reactive carboxyl group on the fatty acid had already been reduced to its alkane form before the rover showed up.

Even more unfortunately, the instrument on Curiosity can only detect straight-chain alkanes that are between 10 and 12 carbons long. So what we see in the Cumberland sample is likely constrained by the limitations of our equipment.

This limitation is particularly vexing because terrestrial life shows a clear preference for even-numbered carbon chains. If we could examine a wider range of alkanes in the Cumberland sample, a relative preponderance of even-numbered alkanes would be strong evidence for their biological origin. But you go to Mars with the instruments you have.

Whatever the exact mix of compounds in the mudstone, it’s their abundance that makes it hard to fit an abiotic origin story to their presence. Curiosity detected the three alkanes at the part-per-billion range. But organic compounds exposed to high-energy cosmic rays degrade over time in a process called radiolysis. And since we know the mudstone in question spent 80 million years being zapped by cosmic rays, we can work backward to estimate the original concentration of medium-chain alkanes or fatty acids in the mudstone when it formed.

Doing this with conservative assumptions, the study authors calculate that the original mudstone contained between 120-7200 parts per million of decane, undecane, and dodecane, or their fatty acid equivalents.

Keep in mind that this is almost certainly an undercount. Since the instrument on Curiosity can’t detect alkane chains outside its narrow window, the total organic content of the sample is probably higher (and note that the 7200 ppm end of the published range is already close to 1% by weight of the entire sample).

Fatty acids are molecules intimately associated with life. They are the constituent parts of triglycerides (fats) and of the lipid bilayer that forms the envelope for every known living cell. There are abiotic pathways to synthesizing fatty acids, but the presence of these molecules (or their alkane cousins) in great abundance on an ancient freshwater lake bed is suggestive.

Like a lot of recent discoveries on Mars, we would have no problem calling this discovery a biosignature if we found it on Earth. Looking back from a distance of 3.5 billion years, this is about as unequivocal as evidence for life gets. There’s sunshine, carbon dioxide, a bucolic freshwater lake, piles and piles of sediment, and abundant organic goo in the sediment with no obvious abiotic source. What more do you need?

But the standard of evidence on Mars is high. So the authors of the paper go to some length to consider alternative sources that don’t implicate biology.

Abiotic hypotheses

Without life, there are four ways you can wind up with a bunch of medium-chain alkanes or fatty acids in an old Martian mudstone:

Read more