The Transit Argument for Space Cities

I want to talk about a billion-person space city. These days, we often use the word space "habitats", which would be the livable volume in an engineered structure. In various cases I'll compare here, a city is a continual habitat, or a cluster of independent habitats that people can move between. The important thing about calling it a "city" is being a connected cluster of living spaces.... and a big one.

We might as well assume the location is somewhere in the Earth-Moon (cislunar) space. To fit the premise, a billion or more should be able to live there, and can commute to another part of the habitat on a daily basis. This distinguishes other visions where a cluster of habitats might exist, but one could not reasonably commute between them daily. That might describe some classic ideas of "flying formations" of rotating habitats.

Why Cities Matter

The real value of a city is not people per square mile. It is jobs, services, and culture (food?) reachable within bounded time. All of those things are complex and messy, but in the big-picture simply scale with the number of people. Thus, the approach I'm going to take is to give the number of people that can be reached within a certain time in transit. All of the other less quantitative sources of value can be said to flow from this.

I've put together some data and scripts to attempt to do this for 6 major cities of Earth, as things stand today. The approach here is a good-enough approximation in order to slap easy numbers on the matter. It takes some real metro stops, with known travel times, and then multiplies out the resultant area by expected population density. This gives the number of people that can be reached traveling to that stop and a time, making 1 data point. Repeat over and over again to get complete data sets.

https://github.com/AlanCoding/gravitational-balloon-mathematics/tree/master/transit/transit_argument

After reviewing the data, I would say that the single least satisfying thing about these results is that it assumes you are downtown in these places. Not only does it assume you're downtown, but that you're basically already inside of the best-available train station. This is not at all how life actually plays out and in Manhattan you're looking at a solid 20 minutes before you even get to that place. All of this is just to explain the unrealistically good numbers. If you add 20 minutes on the top of all of these numbers, it starts to feel more realistic. And that's if you live downtown to begin with! In real life most people are pushed further out from the center, making the curves even worse, by a large margin. Nonetheless, this is still something, and the curves give some satisfying combination of density plus transit capability.

Reachable Population of Earth Metros against Transit Time

This satisfactorily captures current Earth. What can we build in space?

In-Atmosphere Gravity Tubes

For this analysis, I'm assuming 250 meter radius tubes inside of a pressurized vessel in 1km x 1km x 1km lattice. Each tube is home to 1,000 people, making for 1,000,000 tubes in total throughout the volume. I'm using the assumption that people commute from their home tube to some other random tube every morning. They have to ride shuttles and transfer to others. This is done in the microgravity environment through the lattice, and uses different tiered shuttles. These shuttles (like the people) float in open atmosphere, they could use fan control or they could grab onto wires.

• local_float_shuttle: span 1 cell, 1 km hops, 22.2 m/s, 24 people capacity
• regional_express: span 5 cells, 5 km hops, 60 m/s, 192 people capacity
• trunk_axis_line: span 25 cells, 25 km hops, 140 m/s, 1,536 people capacity

Since 1 km is a single lattice, the local shuttle stops at every single tube, which you would expect. The faster shuttles are not unreasonably fast either. The entire structure is 100 x 100 x 100 habitats, so a 25 km hop is going 1/4 the entire edge length. Note that realistically the pressure vessel would be spherical but I'm using simple math here.

The simulation accounts for stopping time and transfer using a constant time for transfers. What does this get you?

Reachable Population of 1 billion Person Space City vs Earth

Same Graph in Log Scale

This is the "wow" factor of the whole idea. You really could transfer within the billion-person city reasonably morning and night. But this undersells it. See reference:

Bettencourt et al., “Growth, innovation, scaling, and the pace of life in cities” (PNAS, 2007)  

The key idea there, which is widely accepted, is that cities are super-linear. As agglomeration happens, the efficiency of delivering increases. Economic output increases faster than population by itself does. So the comparison isn't just 30 million versus 1 billion, but a factor higher than that. More to the point - there is a valid argument that a 1 billion person city will be more compelling to people than Earth cities. This is the transit argument.

If you read the log-scale plot really closely, you might notice that the space city actually reaches less people at small transit time. This is consistent with my assumptions of the city being rural-urban in nature. Specifically, I'm thinking that population density would be on the order of 1,000 square meters per person, or somewhere in that range (most of the current simulation is not exactly specific about this). This is more than comparable Earth metros, so on that factor alone it is expected that the short-time part of the graph favors Earth. Also, due to the lattice spacing itself we have a smaller local density. Basically we are not packing people in really tightly in all the assumptions. You could have assumed otherwise, but this isn't actually a negative. The combination of access + space is something that is actually unique to possibilities in space. That just isn't possible on Earth at all, because of gravity. This is another key reason to believe such space cities will be more compelling to live than on Earth.

In a future post, I hope to go into detail about the alternative where habitats are conventional O'Neill type designs, each having their own atmosphere. So that case requires docking to leave or arrive. But this can still form a 1 billion person city. It's just a little different.