Thursday, November 30, 2023

There Is No Planet B: Part 1

Anything Elon Musk says must be treated skeptically. This is particularly true of anything involving timescales (see Tesla robotaxis). And it is even more true of Musk's plans for visiting and eventually colonizing Mars.

Below the fold in part 1 of this two-part post, I apply some arithmetic just to the logistics of Musk's plans for Mars. Part 2 isn't specific to Musk's plans; I discuss two attempts to list the set of "knowns" about Mars exploration, for which the science is fairly clear but the engineering and the economics don't exist, and the much larger set of "known unknowns", critical aspects requiring robust solutions for which the science, let alone the engineering, doesn't exist:

Amateurs talk strategy, professionals talk logistics
Attributed to General Omar Bradley
Chris Young reported that Elon Musk: SpaceX will build over 1,000 Starships to move 1 million humans to Mars:
The plan is to "build 1000+ Starships to transport life to Mars. Basically, (very) modern Noah's Arks," Musk wrote, reiterating a statement he had made during a recent interview with TED curator Chris Anderson. In that interview, he stated that SpaceX would achieve this goal by 2050.
Wikipedia reports that:
SpaceX and Musk have stated their goal of colonizing Mars to ensure the long-term survival of humanity, with an ambition of sending a thousand Starship spacecraft to Mars during a Mars launch window in a very far future.
SpaceX's Starship is claimed to be able to "carry more than 100 tons of payload to the lunar surface in a single flight" and, using on-orbit refueling, "up to 100 tons all the way to Mars". The Mars mission depends upon refueling the Starship on-orbit from fuel tankers launched beforehand on the Super Heavy booster.

"Musk has predicted that a Starship orbital launch will eventually cost $1 million" but a more realistic estimate is about $1.5M, or $10/Kg. Each launch requires a lot of fuel. "Roughly four hundred truck deliveries are needed for one launch"

The problem with refueling on-orbit is that getting the fuel up for one mission requires a lot of launches. For a manned Moon landing the Government Accountability Office said that SpaceX would "require 16 launches overall". So delivering 100,000 tonnes to Mars would require 16,000 Super Heavy launches costing around $24B and involving 6.4M truck deliveries. One might think that the launches could take the whole of the 780-day interval between Mars launch windows, or a rate of just over 20/day. But because the fuel in the orbital tanks boils off over time, the launches have to happen much faster than that, perhaps around 50/day. in 2022 the Falcon 9 launched 60 times, or 0.16/day. Getting to 50/day requires scaling up over 300 times.

SpaceX's Falcon 9 has an astonishing 99.3% success rate. If the Mars vehicle had the same rate, an extra 113 launches would be needed to cover the failures. Of the 1,000 manned launches, 7 would be failures, carrying 7,000 people.

1000 Starships carrying 1,000,000 people is 1,000 per Starship, which can deliver 100 tons to Mars. That is 220lb/person, so apart from the person and the spacesuit they will need to disembark, they won't have many carry-on bags in the overheads, let alone the air and food needed for the journey. Everything they need to survive on Mars must already have been delivered by earlier missions. Lets guess that each person needs 10 times their weight in life support and other equipment. So the 1M person launch window must be preceded by 10 similar launch windows delivering freight. Now we are talking $264B in launch costs alone, or $264K per person. Even for Elon Musk, over a quarter-trillion dollars is real money. Even that may not be enough. The history of Mars missions shows that Mars landing is a high-risk endeavour. It is very unlikely that all 10,000 freight missions would be successful.

10 launch windows is 7,800 days or more than 21 years. So if Musk wants to land a million humans by 2050 he needs to start launching 50 Starships a day in 2029, a bit more than 5 years from now. Fortunately, he has time to experiment. There are two launch windows before the one when he needs to start in earnest.

If the Starships are to be reused, they have to be re-fueled on Mars using fuel made locally from Martian resources. Each Starship requires 1,200 tons of fuel. Thus unless the Starships are to be expended, during the 2024 and 2026 launch windows Musk needs to deliver a factory to Mars capable of producing 1.2M tons of fuel every 780 days, or nearly 1600 tons/day. Clearly, the factory must weigh several times its daily output. Lets guess it weighs 20 times, or 32,000 tons. So in each of those two launch windows Musk needs to send another 160 Starships to Mars, requiring another 2,560 launches.

But there's another huge problem with Musk's fantasy. The 1,000 Martian emigrants will spend something like 180 days cooped up in each Starship's payload bay, which has a volume of 1,000m3. The average human's volume is around 0.06m3. The cabin volume per passenger of a 1-class 737 is around 0.22m3, so each emigrant will spend 180 days in the equivalent of 4 seats in coach in a 737. I think people paying more than a quarter-million dollars in launch costs alone would be imagining something more like business class! The migrants will end up fighting each other long before they arrive.

And, of course, once the million people land on Mars the story isn't over. They will still be dependent upon supplies from Earth until they can build a completely self-sustaining ecology. They may not need 100,000 tons of supplies every 780 days, but there will still need to be a lot of launches to get fuel into orbit to get a smaller number of Starships with supplies to Mars.

Look, I totally understand that what people like Elon Musk and Jeff Bezos really want is to call up the Magratheans and order one of their custom-made luxury planets lovingly made to their exacting specifications, so they don't have to deal with taxes, government, competitors or people who disagree with them. Let alone having to survive in a 2.5C warmer world racked with war, migration and starvation that is depleting their support staff. It must be really frustrating that niggling little issues like the speed of light mean that the only planet they can afford isn't just a long, slow, expensive commute, but also needs a lot of work to make it a suitable home for a multi-billionaire. That work is the subject of part 2.

All it took for me to get this level of understanding of just the logistics of Musk's fantasy was his own numbers, an Internet connection, a couple of hours, and basic arithmetic. Given Musk's notorious lack of credibility when it comes to schedules, it is disappointing that as far as I can tell no journalist made the effort to inform the public that Musk was BS-ing. Chris Young expressed mild skepticism, "just not very realistic" and "risks putting him in similar territory as he was with Tesla's progress on Level 5 autonomy". But that is a long way from explaining Musk's specific implausibilities.


David. said...

Atrios gets it in Taking Another Big Bong Hit:

"I think we have, at least, passed the stage where journalists printed this shit without qualification."

Followed by a screen-grab of a Tweet from Elon Musk saying:

"We are mapping out a game plan to get a million people to Mars.

Civilization only passes the single-planet Great Filter when Mars can survive even when Earth ship stop coming."

Civilization will fail the single-planet Great Filter long before that because of rich assholes diverting resources from tackling the real problems like CO2 and methane emissions (hello SpaceX) to their fantasies.

David. said...

Eric Berger reports that Elon Musk just gave another Mars speech—this time the vision seems tangible:

"Ultimately, Musk's goal is to seed a civilization on Mars as humanity's first step toward becoming a multiplanetary species. To accomplish this, he believes Mars will need a population of about 1 million people, with many millions of tons of supplies so that settlers can mine and build and grow things on Mars to become self-sustaining.

This will require an absurd amount of launches, 10 per day, and the dispatch of a fleet of hundreds of vehicles to Mars during the short-trip trajectory window that opens between Earth and the red planet every 26 months. Ultimately, while it is challenging, Musk believes that humans could terraform a "fixer-upper" planet like Mars."

See Atrios in the comment above.

David. said...

Musk has updated his timeline for the SpaceX colonization of Mars in a tweet:

"Less than 5 years for uncrewed,
less than 10 to land people,
maybe a city in 20 years,
but for sure in 30,
civilization secured."

So if you believe his schedule it has slipped by 4 years in the past 2 years. To get close to 2054 he needs to be launching 50 Starships/day 4 years after he expects the first uncrewed landing. So the schedule better stop slipping soon.

David. said...

What If We’re Stuck Down Here? by Barry Petchesky discusses yet another reason not to go to Mars:

"The study, ominously titled "Cosmic Kidney Disease" and published last week in Nature Communications, examines the kidney function of 66 astronauts who spent up to 180 days on the International Space Station, which is relatively safe compared to say, a return mission to Mars, which would last a couple years and expose astronauts to the more intense radiation of deep space. But even that limited time really did a number on astronauts. The study found a significant downgrade in kidney function and a higher risk of kidney stones, the result of the shrinkage of renal tubules. This, you need not be a medical doctor to realize, is bad. And the damage could be permanent after long enough—the study simulated the effects of longer exposures on mice, and their kidneys never recovered."