Will Humans Go to Mars before 2033?

[goretexguy]

For reference:
1) Working Starship to/from LEO.
2) Frequent, large-scale cryo propellant xfer.
3) Multi-year, large-scale cryo propellant storage.
4) Comms to/from Mars
5) PNT at Mars
6) EDL proven at Mars
7) Water ice at Mars
ISRU proven at Mars
9) Multi-year life support

This is how I break things down:

Dec 2026 - demonstrate ability to get to Mars, maybe attempt a landing with Optimus. Deploy Marslink satellites for communications and a rudimentary GPS system. (Demonstrates #1 & 4)

Jan 2029 - launch 2 ships a few days apart. Successfully land ship #2 adjacent to ship #1. More Marslinks and Optimus bots. Rollout solar arrays, find ice and begin a fuel generation plant. (Demonstrates #1, 4, 5, 6, 7, & .

2029 (or later) - humans fly extended duration trips on Ships in Cislunar space. Possibly attempt landings on the Moon*. (Assists #5, 6 & 9, possibly 7 if Optimus can wander on the surface and find things.)

2030 (or later) - refuel and launch one of the landed ships on Mars to prove the rocket works after a long idle state and cold soak. Ideally, this vessel would ultimately return to Earth. (At this point, SpaceX should have Marslink receiving stations in Earth orbit, instead of using ground stations which would be subject to all manner of political mischief.) (Demonstrates #1-9)

2031 - launch many (10+) Ships to Mars and land them. These Ships have the basic materials to build a permanent outpost . More Optimus bots to get the work moving along, providing a go/no go for sending people. (Re-demonstration of #1-9)

2033 - launch many Ships to Mars, some with humans, in an audacious, make-Apollo-a-footnote mission.

* Lunar landings don't need humans. Just show that the autopilot can successfully find and land on an unprepared site in non-earthlike conditions.

My 2 cents.

[TheRadicalModerate]

SpaceX is the only substantive and active game in town for H2M.  They’re going to need:

1) Working Starship to/from LEO.
2) Frequent, large-scale cryo propellant xfer.
3) Multi-year, large-scale cryo propellant storage.
4) Comms to/from Mars
5) PNT at Mars
6) EDL proven at Mars
7) Water ice at Mars
8 ) ISRU proven at Mars
9) Multi-year life support

Items 1-5 are needed before 6 can be tested.  Items 1-5 will not be ready by the launch window at the end of 2026, so we’re really looking at the window at the end of 2028 before Item 6 can be tested.

Realistically, SpaceX will need more than one bite at the EDL apple before Item 6 is checked off.  Assuming the late 2028 window gets used up on (and learning from) Mars EDL failures, that means the early 2031 window before work could begin at the Martian surface on items 7-8. 

Like with Mars EDL, realistically, finding Martian water ice is usable form and proving out propellant production will take more than one window.  If the 2031 window gets used up on water ice prospecting or  and propellant production failures and it’s only after the 2033 window that there’s tanks of CH4 waiting for the return leg, then the first crews won’t be sent to Mars until the mid-2035 window.

I think 2035 is a somewhat realistic, median estimate.  I think 2033 and earlier requires SpaceX to get Mars EDL or water ice prospecting/propellant production right within their first windows.  That seems unlikely and unrealistically optimistic.

Personally, I’m even a little more skeptical than that because of the lack of details and apparent work at SpaceX on items 4/5 and 7/8 to date (unless Musk has a secret evil villain volcano base where this work has been going on).  Putting StarLink sats in Mars orbit alone doesn’t provide a link to Earth.  And an Optimus robot walking on the Martian surface isn’t equipped to assay water ice or process propellant.  These are things where NASA experience and expertise could come in really handy so SpaceX doesn’t have to reinvent the wheel.  But with the Trump/Musk fallout, that working relationship will probably not be as close as it has been.  Taking that into account, I don’t think SpaceX will be in a position to send the first crews until the mid-2037 or late-2039 windows.

A decade to a decade-and-a-half still to go is my 2 cents.  FWIW...

The one disagreement I have with your list is that water ice and ISRU prop manufacturing aren't needed.  Assuming the Starship Formerly Known As v3, it's pretty easy¹ to get that landed and returned to LMO using just a single full tank of prop in VLEO.  From there, aerocapturing enough prop to return to Earth is also pretty easy.¹

Even without the SFKAv3, a single v2 only needs about 75t of LCH4 to get to LMO.  If you can land that and a few tonnes of LH2 to catalyze LOX generation using RWGS, that's a far cry from full-up prop production.

_________
¹"Pretty easy" assumes a very, very good cryocooler.

[Robotbeat]

Note that for a given mass of fuel, it is about 100 times easier to actively cool the methane than the hydrogen. In terms of power requirements.

This is due to less heat leakage from a smaller delta-t, less surface area because the propellant is denser, and the far lower temperature differential to pump heat through, which also makes the heat pump both closer to the ideal Carnot efficiency and of course improves the Carnot efficiency.

So if SpaceX pursues active cooling (which they must in order to do ISRU anyway), it’s actually a much easier problem than it would be for Blue origin.

[VSECOTSPE]

It’s also weird to just assume no thought has gone into using Starlinks, in spite of the fact they’re flexible systems, could trade distance for bitrate to keep positive link budget…

This is the kind of handwaving I’m talking about.  Existing Mars spacecraft use X-band for data return.  For high data rates, StarLink is a Ka-band system.  The pointing requirements necessary to use a Ka-band system at Mars distances challenge/exceed the limits of existing systems:

High rate telemetry from Mars requires an extremely large EIRP at Ka-band.  In order to conserve on-board power, extremely large apertures are considered as part of the overall system trade.  Apertures on the order of 1000 wavelengths, at Ka-band, result in antenna bandwidths that approach the attitude knowledge and control capability of modern spacecraft, thereby increasing pointing losses to unacceptable levels.

https://ntrs.nasa.gov/api/citations/20080012561/downloads/20080012561.pdf

Doesn’t mean this and other challenges can’t be overcome.  But unlike we internet space cadets, SpaceX can’t handwave these problems away; they actually have to develop a solution.  My concern or point in the context of estimating initial H2M landing dates is there’s little/no evidence that SpaceX is working these other systems and issues in parallel with Starship.  If that’s true, then addressing these systems and issues serially will add some years to the schedule.

and SpaceX has already used them for talking directly with Dragon and Starship, even in very challenging conditions.

In Earth orbit.  Mars is more than two orders of magnitude farther away.  Challenges like gain and pointing are very different.  Saying that because Starlink works in Earth orbit that it will work at Mars distances is like saying that because my kid can hear me yelling at him from a block down the street, he should be able to hear me yelling at him from 100+ blocks away.  Obviously not true,

It’s not a problem for SpaceX to use the DSN. The DSN is regularly used by other nations, including European spacecraft. This is not a problem. Focus on non-fake problems.

It takes years to make those arrangements with NASA and the DSN.  Unless SpaceX has been working with the DSN on a Mars mission in parallel with Starship, this will add some years to the schedule.  (Hopefully the Trump/Musk divorce doesn’t prevent work towards such an arrangement in the first place.)

It doesn’t necessarily affect the first H2M landing, but the DSN is not a long-term solution for what SpaceX wants to do at Mars.  The DSN is grossly oversubscribed, and incapable of supporting the data demands of hundreds of Mars crew in any event.  SpaceX will need to build its own antenna farm and network infrastructure.

Not if Mars surface cargo logistics has been proven so crews can be resupplied indefinitely and years of supplies are sent ahead of time.

Many contingencies can’t be solved with supplies.  There are reasons why crews and crewmembers would need to return to Earth that have nothing to do with running short on supplies.

When you have a hammer, like Starship’s large payload mass, everything starts to look like a nail.  But many problems are screws and bolts, and you still need some screwdrivers and wrenches, like a crew return capabilty, to address those other contingencies.

[VSECOTSPE]

The one disagreement I have with your list is that water ice and ISRU prop manufacturing aren't needed.  Assuming the Starship Formerly Known As v3, it's pretty easy¹ to get that landed and returned to LMO using just a single full tank of prop in VLEO.  From there, aerocapturing enough prop to return to Earth is also pretty easy.¹

If I understand you correctly, we’re trading the complexity of prospecting for water and manufacturing propellant on the Martian surface for an RPO in Mars orbit between returning Starships and some flavor of a depot Starship sent to Mars orbit.  Once you’ve proven RPO in Mars orbit, that simplifies a lot operationally.  But in terms of reducing the time to the first H2M landing, I don’t know that it saves us anything.  Propellant production at Mars and RPO in Mars orbit are both unproven capabilities.  I’d guesstimate that both impose a high likelihood of initial failure/steep learning curve that will eat up a launch window or two solving the same bottleneck before a crew can be sent.

Schedule aside, the handwaving in Musk’s presentations about finding water and ISRU is concerning.  Even after a couple decades of missions and research, planetary scientists still debate whether the geological signatures (like “weepings”) that we associate with water at Mars are actually due to water.  SpaceX needs some backup solutions like forwarding depots to Mars orbit.

Even without the SFKAv3, a single v2 only needs about 75t of LCH4 to get to LMO.  If you can land that and a few tonnes of LH2 to catalyze LOX generation using RWGS, that's a far cry from full-up prop production.

Obviates the water prospecting, which is good.  But no one has done the Sabatier process at Mars and getting all the details of maintaining the correct temperatures and pressures for the reaction and getting the impurities separated, the gases separated, and the reverse water gas shift implemented in a foreign environment may still consume a launch window or two.

Professionally, I lived through a couple decades of STS launch delays due to hydrogen leaks, so long-term LH2 storage makes me nervous, both from an operational and safety standpoint.  But being able to forgo water prospecting is obviously worth the risk.

[TheRadicalModerate]

If I understand you correctly, we’re trading the complexity of prospecting for water and manufacturing propellant on the Martian surface for an RPO in Mars orbit between returning Starships and some flavor of a depot Starship sent to Mars orbit.  Once you’ve proven RPO in Mars orbit, that simplifies a lot operationally.  But in terms of reducing the time to the first H2M landing, I don’t know that it saves us anything.  Propellant production at Mars and RPO in Mars orbit are both unproven capabilities.  I’d guesstimate that both impose a high likelihood of initial failure/steep learning curve that will eat up a launch window or two solving the same bottleneck before a crew can be sent.

RPO would likely require a fairly robust PNT system.  But I think that some Marslink deployment would go a long way to retiring RPO risk.¹

Schedule aside, the handwaving in Musk’s presentations about finding water and ISRU is concerning.  Even after a couple decades of missions and research, planetary scientists still debate whether the geological signatures (like “weepings”) that we associate with water at Mars are actually due to water.  SpaceX needs some backup solutions like forwarding depots to Mars orbit.

I don't think the presence of water is a requirement for a human Mars mission.  It is a requirement for colony-scale (and possibly base-scale) settlement.  But early exploration missions need to be able to keep their crews alive--and return them alive--irrespective of the presence of ISRU water.  IMO, that requires full importation of prop, or at least importation of LCH4 or LH2 (to make LCH4).

Obviates the water prospecting, which is good.  But no one has done the Sabatier process at Mars and getting all the details of maintaining the correct temperatures and pressures for the reaction and getting the impurities separated, the gases separated, and the reverse water gas shift implemented in a foreign environment may still consume a launch window or two.

I'm not proposing Sabatier here.  It's easy to land enough LCH4 to at least get back to LMO.  I believe it's not much more difficult to land enough to get the crew all the way back to Earth.

The question is how to make LOX.  At least two possibilities:

1) MOXIE-style CO2 electrolysis.  Currently very energy intensive, and very low-scale.

2) Reverse water-gas shift (RWGS).  There's a huge amount of industrial experience with WGS, and a fair amount with RWGS:  CO2 + H2 --> CO + H2O.  Vent the CO, electrolyze the water, cryocool the O2,  and recycle the hydrogen.

Professionally, I lived through a couple decades of STS launch delays due to hydrogen leaks, so long-term LH2 storage makes me nervous, both from an operational and safety standpoint.  But being able to forgo water prospecting is obviously worth the risk.

My intuition is that the less LH2 (or maybe just compressed GH2) you need, the better.  A lot depends on how efficient you can make your RWGS rig.  If you can make it really efficient, you might easily be able to make enough LOX for a return mission with only a couple of tonnes of LH2.  That allows you to put it in a proper dewar, with proper insulation and heavy-duty fittings, to eliminate its tendency to jailbreak.

So the real question is how well you can scavenge the water from your RWGS setup.

________
¹You had objections to Marslink hand-wavery, which are legitimate.  However, I'm considerably less worried about pointing accuracy then you are, because you can have the entire constellation relay interplanetary traffic through a big honkin' bird that has the proper antennae at the proper frequencies.

However, I don't think that RPO requires any interplanetary traffic.  It mostly requires PNT, which remains local to Mars.

Interplanetary bandwidth is an issue.  But it's less of an issue if you have high bandwidth in the local environment, and lots of memory.

[Robotbeat]

CO2 electrolysis isn’t terribly challenging. MOXIE was a small scale demonstration, but solid oxide electrolysis has been done at a pretty large scale and so scaling up MOXIE is very straightforward and easy to test on Earth. The efficiency is just as good as RWGS.

[Vultur]

7 and 8 don’t have to be figured out before they go.

Before the first crews go, the program needs tanks of LCH4 on the Martian surface for crew return options. 

I don't think Elon Musk necessarily agrees with that.

I do *not* think contingency return is expected to be an option immediately. If something goes wrong with ISRU, crew may have to wait for the next synod cargo delivery.

(And I'm OK with that. Given travel times, you can't get back to Earth *quickly* anyway. You need to be able to deal with emergencies with what you bring.)

OTOH, I also think the initial crew will be accepting risks far higher than NASA would be comfortable with.

[TheRadicalModerate]

CO2 electrolysis isn’t terribly challenging. MOXIE was a small scale demonstration, but solid oxide electrolysis has been done at a pretty large scale and so scaling up MOXIE is very straightforward and easy to test on Earth. The efficiency is just as good as RWGS.

Let's figure this out, as much as possible.  Wikipedia says that CO2 electrolysis needs 25-30kW to produce 2kg/hour.  That's about 13kWh/kg, right?

The back-end of the RWGS process requires water electrolysis, which is 55kWh/kg of H2, which would be 6.9kWh/kg of O2.  I don't have a clue how to estimate energy needed to produce H2O through RWGS, but if it's more than about 5.4kWh/kg of H2O, then CO2 electrolysis has it beat.  Since the reverse reaction is endothermic, it's probably not great.

However, both of these are kind of appalling.

I can get a v2-ish Starship back to Earth orbit (propulsive capture, but it'll need refueling), for 900t of LOX.  To produce that in two years (not absolutely necessary; it could be 4.2 years, or 6.4 years), you'd need to produce 1233kg/day.  Scaling, that would require 15.4MW of firm power.

Using that model I built for our APU vs. solar argument (solar won handily for Mars), that's a power system mass of 1940t.  I can land all the prop I need (1150t of methalox) for considerably less.¹

That doesn't mean that it's silly to produce ISRU prop; it just means that it's expensive to start up.  For the earliest missions, it almost certainly makes more sense to land the prop, or land enough prop to get to LMO to meet a depot.

_______
¹This is all prop landed and propulsive capture, which retires most of the risk that VSECOTSPE was objecting to. 

You can do substantially better than this by landing just enough prop to get back to LMO, and a depot in orbit for the rest.  And if somebody will actually let you aerocapture/EDL (I don't think that's going to happen, but who knows?), you can do even better.

[Robotbeat]

Starship requires an enormous amount of energy to make propellant.

It’s best to just look at the high heating value of the fuel and multiply by 2 (50% efficiency) to get the required input.

[VSECOTSPE]

I don't think Elon Musk necessarily agrees with that.

I do *not* think contingency return is expected to be an option immediately. If something goes wrong with ISRU, crew may have to wait for the next synod cargo delivery.

(And I'm OK with that. Given travel times, you can't get back to Earth *quickly* anyway. You need to be able to deal with emergencies with what you bring.)

OTOH, I also think the initial crew will be accepting risks far higher than NASA would be comfortable with.

A lot of this comes down to how hard is it to establish that Starship return capability.

If it’s as simple as sending some extra Starships one-way with extra LCH4 tankage, then why not establish that Starship return capability before the first crews go?

If it’s as complex as producing LCH4 locally from Martian water ice and atmosphere, then SpaceX may not even be able to do it remotely in the first place and have to send a crew to establish that initial Starship return capability anyway

Reality will probably be somewhere in between.

There’s not enough detail in Musk’s presentations to know exactly what SpaceX has planned.  ISRU/propellant production comes the synod after the first human landings in his timeline.  But whether Starship return is dependent on ISRU/propellant production is unclear.  Starship payload capability may determine whether establishing Starship return from Mars is easy or hard, but Starship capability is a moving target for now, so even SpaceX may not know internally yet.  (Or SpaceX just has not dealt with these details while Starship consumes all the talent and man-hours.)

Like so many things, flight safety is too often argued in terms of poles.  Either no price is too high to ensure astronaut lives or any effort to improve safety is a wasteful distraction from the mission.  In reality, resources are always constrained, and there is no bigger distraction from the mission than dead or dying astronauts.  So it’s a question of best effort.  Is the organization/leadership/program doing everything reasonable to ensure crew health and safety?  Are they wasting large amounts of resources polishing a safety issue for little return?  Or are they ignoring modest investments that could dramatically improve crew survivability?  At OMB, I used to review laundry lists of potential Shuttle upgrades through these exact lenses.

My sense is that a Starship return capability is something that could be reasonably afforded and established within a synod or two that would inoculate the program against a range of scenarios that could consume the program in its early, tender years.  The crew/program/leadership/stakeholders/public doesn’t want to an astronaut with, say, a probable cancer diagnosis sitting on the surface of Mars with no way back to Earth conjunction after conjunction at the same time it’s trying to get a base camp established. 

I’d have a different recommendation if I knew it would take five synods to establish that Starship return capability.  But in general, I favor modest delays to increase sustainability and the probability of long-term success over meeting artificial deadlines.  I’d like to think SpaceX will move away from the Apollo-end-of-the-decade and Artemis-beat-the-Chinese-by-2030 mentalities that don’t lead to anything sustainable.

My 2 cents... YMMV.

[lykos]

You have to change the art of thinking.
Don't expect and don't plan to bring the astronauts back in first place.
When it's clear from the beginning, that they have to stay at least for one synode, many problems are solved.
There are no expectations and no disappointments.
There are big risks and people will die that's a fact. You can try to make it save, and you have to, but people will die.
Do not make to big promises.

[sdsds]

If a crew of ten departed Earth in 2031 and one of them were to die from cancer sometime after arrival we would mourn the loss, and honor the deceased as the first human to die on Mars. If all in the crew were to die because the resupply ship didn't arrive in 2033, that would be shameful. Not for them, but for those of us still on Earth.

[TheRadicalModerate]

You have to change the art of thinking.
Don't expect and don't plan to bring the astronauts back in first place.
When it's clear from the beginning, that they have to stay at least for one synode, many problems are solved.
There are no expectations and no disappointments.
There are big risks and people will die that's a fact. You can try to make it save, and you have to, but people will die.
Do not make to big promises.

The issue isn't if some people die.  But if they all die, that's another story.

I think a single-synod nominal mission is a desirable baseline.  The big question is whether it's worth planning for an opposition-class short-stay abort.  They don't exist for all windows, and sometimes their outcomes may be riskier than hunkering down and waiting for the next conjunction.

My guess is that, by the time they get to Mars, a crewed mission will have uncovered multiple instances of systems that are underperforming or degrading faster than expected.  If mission management has an abort in its toolkit, some of those contingencies might trigger it.  If an abort doesn't exist and the systems are going to fail before the next window opens, that's a problem.

This all seems like Mission Planning 101 to me:  Think about what can go wrong, provide spares and backups when possible, see if there are other ways to mitigate the contingencies, and be as ready as possible for the unexpected. 

Above all, don't be reckless.  Taking informed risks is just part of human spaceflight.  But the emphasis should be on the "informed" part.

[Vultur]

My guess is that, by the time they get to Mars, a crewed mission will have uncovered multiple instances of systems that are underperforming or degrading faster than expected.

Why? I would *not* expect that - before getting to Mars, the systems won't be experiencing conditions all that different from what they'd have experienced on previous crewed test flights. (I don't expect them to make the Artemis II mistake of never testing the life support system with actual humans before a deep space mission.)

[sdsds]

In their Table 4, M. Vasile et al. / Acta Astronautica 56 (2005) 705 – 720 present an interesting free-return trajectory with a launch date in 2032.
https://www.esa.int/gsp/ACT/doc/MAD/pub/ACT-RPR-MAD-2005-%28Acta%29DesignOfEarthMarsTransfersWithEvolutionary.pdf

Δv1 and Δv4 are the Earth departure and return maneuvers. This seems ... not impossible.

[Helodriver]

It still surprises me to see the degree of handwavium being consumed about orbital refueling, loitering depots, interplanetary cruise and surface operations on Mars where storage of cryogenics is concerned.

The longest duration cryogenics have ever been used in American space flight is 17.5 days on a shuttle with an Extended Duration Orbiter tankage kit. Thats it. EDO was just an array of well insulated spherical tanks plumbed into the existing orbiter cryogenic system to power fuel cells and were not refilled, exposed to off nominal solar heating, or EDL stress. Starship needs its thin walled orders of magnitude larger tanks to contain two kinds of cryo for months or years without leaking, excess boiloff or breakdown of cryocoolers, fittings and seals, an unprecedented challenge without analogue. Its loss of mission and/or vehicle and crew if any of that fails and is likely to be far more difficult if not the most difficult technical task, up there with EDL on unprepared surfaces.  It may take multiple synods just to get that right, notwithstanding other issues.

[TheRadicalModerate]

It still surprises me to see the degree of handwavium being consumed about orbital refueling, loitering depots, interplanetary cruise and surface operations on Mars where storage of cryogenics is concerned.

The longest duration cryogenics have ever been used in American space flight is 17.5 days on a shuttle with an Extended Duration Orbiter tankage kit. Thats it. EDO was just an array of well insulated spherical tanks plumbed into the existing orbiter cryogenic system to power fuel cells and were not refilled, exposed to off nominal solar heating, or EDL stress. Starship needs its thin walled orders of magnitude larger tanks to contain two kinds of cryo for months or years without leaking, excess boiloff or breakdown of cryocoolers, fittings and seals, an unprecedented challenge without analogue. Its loss of mission and/or vehicle and crew if any of that fails and is likely to be far more difficult if not the most difficult technical task, up there with EDL on unprepared surfaces.  It may take multiple synods just to get that right, notwithstanding other issues.

ZBO doesn’t take several synods. It may take several multi-week missions, but they can happen in cislunar space or even LEO.   Mars EDL, on the other hand, is quantized in synod increments, because the only place it can be tested is on Mars.

[Robotbeat]

Mars EDL is similar to Earth EDL. In fact, NASA often tests parachutes or new heatshields (HIAD) or whatever at Earth. Sometimes that’s too expensive, like with sky crane, but it is among the options NASA could use. It’s not a perfect copy, but the differences can be accounted for with arcjet and wind tunnel tests.

Of course, SpaceX is going to get PLENTY of practice doing Earthside EDL. And the more tanker trips required, the more EDL practice.

This is one reason I sort of wish SpaceX would do more legged landing tests with Starship instead of just chopsticks. I suppose they may still do some since it’s needed for Earth 2 Earth rocket cargo to austere locations, ie the Air Force contract. Of course, they do have lots of experience doing landing tests to prepared flat landing pads with Falcon (literally hundreds of tests), and starship did do a few suborbital landing tests with legs, but trying to find a landing site on Mars that just happens to be equivalently flat seems unlikely. So I hope they do do some non-flat landing tests before Mars 2026. (Although in the latest presentation, no legs are shown for the 2026 mission but are for the later ones, implying they may attempt to land on the skirt or use the tiny flip out legs… neither option seems reliable for an unprepared site, to say the least.)

This is one thing that lunar missions would actually help with.

[Robotbeat]

Helodriver: it’s totally false that the shuttle extended orbiter kit is the longest duration cryogenics on an American spaceflight mission. Several space telescopes or other missions operated for months or years using either a cryogenic tank or a combination of passive and active cooling (JWST).