What does it take to get to a million people on Mars?

[john smith 19]

Elon Musk has said he wants to build a city on Mars to have a million people in it.

The BFS is designed to be the biggest spaceship that has ever left Earth's orbit but how many of them would be needed to do this in a reasonable amount of time?

To get some idea of the options to this, and the time scales they would give, I have built a little simulation game in Excel.  It's split into 2 windows, so you can change the parameters and see the growth patterns of various parameters. As usual cells with Blue borders are for user entry. All others are generated from them.

The game lets you change assumptions on the fleet of BFS's going to Mars and the passengers it will carry. The  fleet calculation supports a "learning" curve factor that can lower the cost of BFS's as the size of the fleet expands.

Population growth is much more tricky to describe over time  as AID and IVF radically open up the options for childbirth. The game uses a simple idea that people on each new arriving fleet pair up and have a certain average number of children shortly after arrival. 10 launch windows (that's hard wired in, despite it being possible to shorten the gaps between launch windows) later these children then have an average number of children also.

The game runs up to 52 launch windows, or 104 years from the landing of the first ships crewed ships. 
The sheet shows wheather the assumptions result in the settlement exceeding a million people and if so what year that happens. An interesting range to aim for would be around launch window 24 or 48 years. Assuming the first landing is 2022 that would be when Musk is 99 years old.

I'll let people download it and play with it first before commenting on some of the results.

[QuantumG]

He also said the later ships would make the ITS (the bigger design before the BFS) look small. So it's not really sensible to talk about doing it with BFS.

[john smith 19]

He also said the later ships would make the ITS (the bigger design before the BFS) look small. So it's not really sensible to talk about doing it with BFS.

The number of ships built before each launch window, and the number of passengers they carry, can both be varied. To a first approximation any potential "BFS 2.0" can be treated as an increase on the average number of passengers and a reduction in the number of vehicles built (along with an increase in the first cost of the ships, which can also be changed).

But I'll keep that in mind if there's enough interest in doing an upgrade for the game.

[meekGee]

He also said the later ships would make the ITS (the bigger design before the BFS) look small. So it's not really sensible to talk about doing it with BFS.

Also, population growth.  Considering a 50 yr timeline, people will do what people do.

[Lar]

He also said the later ships would make the ITS (the bigger design before the BFS) look small. So it's not really sensible to talk about doing it with BFS.

Also, population growth.  Considering a 50 yr timeline, people will do what people do.

He said you can tweak the average number of children per pair (But not how long it takes to get them)

[meekGee]

He also said the later ships would make the ITS (the bigger design before the BFS) look small. So it's not really sensible to talk about doing it with BFS.

Also, population growth.  Considering a 50 yr timeline, people will do what people do.

He said you can tweak the average number of children per pair (But not how long it takes to get them)

Yup.

For a colony, it will be a cultural thing...  Assume n>>2...  Look at any frontier society.  N=6?

A generation is ~25 years.

So over 50-100 years, self-growth will be a big factor even if there's no influx.

Which is good, since it's hard to move 1000000 people when you can only go every 2 years...

[Robotbeat]

5 engines per BFS. They currently can make like 200 engines per year, let's say they up that to 400. That's enough, over a 30-32 year BFS lifespan, to keep over 1000 BFSes in circulation. Up to 200 people per BFS (stretched a bit, passengers only, 5 people per cabin), that means they can move 200,000 people per synod. Only takes ~10 years to move a million people. Longer if you assume only 2-3 people per cabin.

It's not actually absurd to talk about moving a million people with BFR/BFS.

[jpo234]

5 engines per BFS. They currently can make like 200 engines per year, let's say they up that to 400. That's enough, over a 30-32 year BFS lifespan, to keep over 1000 BFSes in circulation. Up to 200 people per BFS (stretched a bit, passengers only, 5 people per cabin), that means they can move 200,000 people per synod. Only takes ~10 years to move a million people. Longer if you assume only 2-3 people per cabin.

It's not actually absurd to talk about moving a million people with BFR/BFS.

Elon also said, that he expects that for every passenger flight there will be 10 cargo flights.

[meekGee]

5 engines per BFS. They currently can make like 200 engines per year, let's say they up that to 400. That's enough, over a 30-32 year BFS lifespan, to keep over 1000 BFSes in circulation. Up to 200 people per BFS (stretched a bit, passengers only, 5 people per cabin), that means they can move 200,000 people per synod. Only takes ~10 years to move a million people. Longer if you assume only 2-3 people per cabin.

It's not actually absurd to talk about moving a million people with BFR/BFS.

Elon also said, that he expects that for every passenger flight there will be 10 cargo flights.

That can only apply in early stages...

[john smith 19]

He also said the later ships would make the ITS (the bigger design before the BFS) look small. So it's not really sensible to talk about doing it with BFS.

Also, population growth.  Considering a 50 yr timeline, people will do what people do.

He said you can tweak the average number of children per pair (But not how long it takes to get them)

Shortening the pregnancy cycle for women is a whole other thing. 

My instinct is that somehow ways will be found to shorten the gap between launch windows. America was settled with 8-12 week voyages but with (in principal) daily departures in both directions. Sets of BFS's on cycler orbits (1 fast out, multiple slow "outs" to give the equivalent fast returns later on)?, ion thrusters after initial boost? (Dare I suggest it?) Nuclear thermal? Within the SoA but ruinously expensive to design and test 9and unlike Kilopower I don't think NASA has any actual hardware in test. Yet)

I know it will take a lot of delta V to do this but let's remember compared to nearly everything else in spaceflight propellant (anything, including LH2, except NTO/Hydrazines) is cheap.

5 engines per BFS. They currently can make like 200 engines per year, let's say they up that to 400. That's enough, over a 30-32 year BFS lifespan, to keep over 1000 BFSes in circulation. Up to 200 people per BFS (stretched a bit, passengers only, 5 people per cabin), that means they can move 200,000 people per synod. Only takes ~10 years to move a million people. Longer if you assume only 2-3 people per cabin.

It's not actually absurd to talk about moving a million people with BFR/BFS.

No one said it wasn't. It just needs enough of them.
Prior to each launch window you are planning to add another 80 BFS's to the fleet and they will carry 200, not 100 passengers. Being a bit crude this also assumes the first 2 will seat 200 as well, but it gives you the idea.

And you could send more than a million people to Mars in 24 years, which is well below "retirement age" for BFS (good point. I'll have to add an "age out" entry to the set up). That excludes any children to increase the total. An average of one child per couple would top a million in less than 20 years.

The cost to build that fleet would be $93294m IE about $93.3Bn. 

Starlink will have to sell a really large amount of bandwidth to build that fleet, although 200 passengers at $500K is $100m. Now will that offset only the propellant or some of the build cost as well?

But at 6 children per couple (Meekgee's preferred level) that's only 14 years and $71.5Bn in build costs.

Yup.

For a colony, it will be a cultural thing...  Assume n>>2...  Look at any frontier society.  N=6?

A generation is ~25 years.

The number in the game is an average. Keep in mind if someone had only a couple of children that would mean another couple would have 4 more, or 2 couples would have another 2 each, or 4 couples would have another 1.

I'd suggest there is a certain amount of cultural variation on what constitutes a generation.

So over 50-100 years, self-growth will be a big factor even if there's no influx.

Provided there are no major issues with gestation and child birth in 1/3 g. You might like to look up the evidence for that.

Which is good, since it's hard to move 1000000 people when you can only go every 2 years...

Hard, but not impossible. Just very expensive.

Elon also said, that he expects that for every passenger flight there will be 10 cargo flights.

Good point. Another one for the to-do list.
"Percentage of flights that will be cargo only"

Looks like there will be a V 2.0 after all.

[john smith 19]

That can only apply in early stages...

Should only apply in early stages.

But it depends.

As you move to locally made products the range of goods you need to buy in narrows, but the number of people who need those bought in goods has grown.  If the former exceeds the latter then then the overall proportion of cargo flights falls. If not it stays the same, despite the settlements greater self sufficiency.

[launchwatcher]

Starlink will have to sell a really large amount of bandwidth to build that fleet, although 200 passengers at $500K is $100m. Now will that offset only the propellant or some of the build cost as well?

Raptor propellant is cheap.

IAC2017 slides say total BFR takeoff mass is 4400T.

Assume it's all propellant (we're just going for a ballpark estimate here..)

Assume a mass ratio of 1kg CH4 to 3.6kg LOX; 4400/4.6 = 957 so:

957t CH4
3443t LOX

Once source I found said NASA paid $0.16/kg for LOX.   That's $160/t.   

Natural gas is mostly methane, and the price of natural gas is dynamic.   Let's pick one of the higher prices (~$10 per thousand cubic feet) which is ~$450/t

so one BFR-load is 957*450+3443*160 ~= $980,000.   

So each pair of $500k tickets buys the raw fuel for one BFR launch.   If it takes 6 BFR launches to get one BFS to Mars, you need to sell 12 $500k tickets (or 24 $250k tickets)  to pay for the fuel.

[Hotblack Desiato]

In theory, they should be able to ramp up the production rate of BFS aswell. Means, from these initial 8 BFS to something more reasonable, like 20 or even more per launch window, and from there to 100, etc... (essentially exponential growth of production and launch capacity).

And for financing BFS, we should factor in a Moon Colony and LEO-stations. BFS can earn additional money by doing these short trips over there. If a BFS lasts 25 years and/or 100 launch-reentry cycles, that would waste 76 cycles because it just can't do more than 12 flights to Mars and back (especially since landing on Moon doesn't count as atmospheric reentry).

Another intersting point will be the price tag on the ticket. $ 500k is a nice value, but Musk already indicated that it might be much lower, in the ballpark 200k or even 100k. For Mars, it seems to be 6 launches to get a full set of 100 passengers to Mars, while Moon requires just 2 launches and LEO even one launch (maybe Moon and LEO can be done with more than 100 passengers, just because its a much shorter time period for the flight (3-5 days vs 120 days - less consumables, less internal volume requirements, etc). Means a Moon ticket could be at $20-100k and a LEO ticket at below 10-20k. And all that helps keeping the price low for a ticket to Mars.

I think, we won't see that 1 million number within the first 25-30 years (first flights will just have a very limited number of passengers) but once the regular BFS flights with >100 passengers start happening, we will see numbers rising pretty quickly. Once that million is reached, it won't stop. For example, in your file, with your standard-settings, we have the million people at window 50. but already at window 61, it's 1.5 million and at 70 it's 2 million. 
(and at this point, we will exceed 10 million on the Moon, just as a sideproduct of the Martian Colonization effort)

Will be very interesting to see the future.

[john smith 19]

Starlink will have to sell a really large amount of bandwidth to build that fleet, although 200 passengers at $500K is $100m. Now will that offset only the propellant or some of the build cost as well?

Raptor propellant is cheap.

IAC2017 slides say total BFR takeoff mass is 4400T.

Assume it's all propellant (we're just going for a ballpark estimate here..)

Assume a mass ratio of 1kg CH4 to 3.6kg LOX; 4400/4.6 = 957 so:

957t CH4
3443t LOX

Once source I found said NASA paid $0.16/kg for LOX.   That's $160/t.   

Natural gas is mostly methane, and the price of natural gas is dynamic.   Let's pick one of the higher prices (~$10 per thousand cubic feet) which is ~$450/t

so one BFR-load is 957*450+3443*160 ~= $980,000.   

So each pair of $500k tickets buys the raw fuel for one BFR launch.   If it takes 6 BFR launches to get one BFS to Mars, you need to sell 12 $500k tickets (or 24 $250k tickets)  to pay for the fuel.

Thanks for that. I couldn't find my download of Musks 2017 IAC presentation.  That also implies there would be money left over to offset BFS construction costs. I went with an initial cost of $400m but the sheet includes a learning curve to lower costs as production experience grows, although how much by is debatable.

[john smith 19]

In theory, they should be able to ramp up the production rate of BFS aswell. Means, from these initial 8 BFS to something more reasonable, like 20 or even more per launch window, and from there to 100, etc... (essentially exponential growth of production and launch capacity).

True, if the existing facilities can accommodate the ramp up, otherwise you're looking at serious upgrade costs.

Another intersting point will be the price tag on the ticket. $ 500k is a nice value, but Musk already indicated that it might be much lower, in the ballpark 200k or even 100k. For Mars, it seems to be 6 launches to get a full set of 100 passengers to Mars, while Moon requires just 2 launches and LEO even one launch (maybe Moon and LEO can be done with more than 100 passengers, just because its a much shorter time period for the flight (3-5 days vs 120 days - less consumables, less internal volume requirements, etc). Means a Moon ticket could be at $20-100k and a LEO ticket at below 10-20k. And all that helps keeping the price low for a ticket to Mars.

In the 2016 presentation Musk said he expected a Mars ticket would cost the equivalent of "The mediun price of a house in the United States." Today that is about $189 000.

I think, we won't see that 1 million number within the first 25-30 years (first flights will just have a very limited number of passengers) but once the regular BFS flights with >100 passengers start happening, we will see numbers rising pretty quickly.

Actually it is possible. The whole point of the game is to see what you have to do if you want it to happen.

For example with an average 100 passengers a ship (to overcome the early flights which were not full up) you could do it with a fairly modest 2 children per couple  in 24 years if you're ready to build 80 new BFS's per launch window. 

My instinct for BFS is that as SX gain experience and become comfortable with ECLSS margins they will carry more people. I think they will build more BFS's to add to the fleet with every launch window, but more like 8 than 80 at a go. Time will tell who's right.

 IIRC the pads in Florida were designed to support up to 14 million lbs of thrust. The hard limit is set by this.  Beyond that SX will be plowing up cubic miles of concrete to go bigger.

So the first upgrades will be whatever the BFS ECLSS can actually carry (20more? 30more?)  and "BFS 2" will be up to the full T/O thrust limit of the pads, about another 14% bigger (5500 tonnes of thrust is already about 12.1 million lbs).

My bet is on SX (somehow) spending more propellant to widen the launch window. If SX have taught one lesson it is propellant is (relatively) cheap but time is expensive. Even cutting the time between launch windows a few months makes a huge difference in the long term. [EDIT. For example the launching of the fleet can be broken up into smaller units (call it a "squadron" for example), so less of a "surge" and more of an even work flow as each group departs. This also offers the chance for any last minute orders for parts or equipment to be loaded onto the last group to leave. ]

[JamesH65]

I don't think the limit is how fast you can get them there, but how fast you can build somewhere for them to live.

How many cities on Earth, built from scratch, had 1M inhabitants in less than 50 years, less than 100 years etc in STABLE accommodation (i.e. not shanty towns). Not many, if at all.

[RonM]

BFS is a good architecture to get Mars settlement started, but not practical for large scale.

Once there is enough infrastructure on Mars to allow for rapid settlement expansion, a larger scale system can be deployed. Big spacecraft to transfer settlers by the thousands will travel back and forth between Earth and Mars. These spacecraft will be refueled in Earth orbit using propellant created from resources mined from the Moon and NEAs. Once they arrive in Mars orbit, they can be refueled from Mars, Phobos, Deimos, or asteroid mined resources.

Passengers will be transported to and from orbit by smaller vehicles, such as BFS.

Blue Origin, ULA, and others can develop cislunar space to proved the Earth end of this infrastructure for SpaceX to buy propellent.

[LM13]

He also said the later ships would make the ITS (the bigger design before the BFS) look small. So it's not really sensible to talk about doing it with BFS.

Also, population growth.  Considering a 50 yr timeline, people will do what people do.

He said you can tweak the average number of children per pair (But not how long it takes to get them)

Yup.

For a colony, it will be a cultural thing...  Assume n>>2...  Look at any frontier society.  N=6?

A generation is ~25 years.

So over 50-100 years, self-growth will be a big factor even if there's no influx.

Which is good, since it's hard to move 1000000 people when you can only go every 2 years...

I wouldn't be so sure about N=6.  I'd expect N=3 for a while to be the high end, with most couples closer to just 2.  The reason is that education tends to correlate with reduced birth rates (something Musk himself has complained about, remarking in the Vance biography that not enough intelligent women are having enough children).  The kind of people most likely to go to Mars for a while will generally be more educated, which also makes them older when they get started on the trip, and thus less fertile (30 y.o. people vs. 20 y.o. people).  The likely absence of specialized medical care for special-needs children (more likely for older parents) will also likely prompt caution in procreation for these older citizens. 

Unless, of course, we propose the existence of specialized schools that optimize their graduates for life on Mars by the age of 18, and either a "Mars Loan" program to be repaid over time, or a scholarship/trust fund. 

[john smith 19]

Once there is enough infrastructure on Mars to allow for rapid settlement expansion, a larger scale system can be deployed. Big spacecraft to transfer settlers by the thousands will travel back and forth between Earth and Mars. These spacecraft will be refueled in Earth orbit using propellant created from resources mined from the Moon and NEAs. Once they arrive in Mars orbit, they can be refueled from Mars, Phobos, Deimos, or asteroid mined resources.

Passengers will be transported to and from orbit by smaller vehicles, such as BFS.

That would certainly save the wear and tear on the vehicles that have the fewest opportunities to amortize their costs (IE the inter orbit transports).

Obviously getting to a stage where the budgets for such an optimized vehicle is justified (and you can bet that won't be cheap) is going to take some time.

[john smith 19]

I don't think the limit is how fast you can get them there, but how fast you can build somewhere for them to live.

How many cities on Earth, built from scratch, had 1M inhabitants in less than 50 years, less than 100 years etc in STABLE accommodation (i.e. not shanty towns). Not many, if at all.

A fair point but keep in mind the next population burst is very predictable (barring really massive breakthroughs in propulsion), so the settlement has 2 years to get the last batch "settled" before the next batch arrive.

Of course the population growth model does assume half of these people will be getting pregnant so thre will be a lot of construction going on to build out the settlement.