The Space Reviewin association with SpaceNews

L5 space colony illustration
Space colonies located in L5 were all the rage in the 1970s, but today a more incremental approach starting on the Moon may make more sense. (credit: NASA Ames Research Center)

The high risk frontier

In The High Frontier: Human Colonies in Space, by Gerard O’Neill and published in 1977, a very exciting case is made to take humanity into space to stay. With plentiful and reliable sunlight, good access to the asteroids, Earth, and the Moon, L5 has a bright future. With almost 30 years of hindsight, much of the utopian vision has been validated even as some of the technologies remain 30 years away (see sidebar).

The High Frontier tells us more about the 1970s than it does about human space colonies. The social, political, and economic case to emigrate is nearly identical to the case made in the Club of Rome report that says that we are going to run out of everything by the year 2000. O’Neill waxes poetic about the cost and capability of the latest, greatest lifter, the space shuttle. An orbiting L5 station is offered as a cure for inflation. Even the easier attitude toward sex is present: “can one imagine a better location for a honeymoon hotel than the zero-gravity region of a space community?” I am much more controversial today even raising the subject, much less saying that honeymooning couples would probably prefer the attraction of gravity.

The High Frontier tells us more about the 1970s than it does about human space colonies.

The technical and spiritual cases are more timeless. We want to become a spacefaring species. L5 and the Moon offer two great spots for industrial development that can support development of the asteroids, the outer moons, and the rest of the solar system. A frontier even for a small number of people can spur the technological challenge of solving age-old labor productivity problems since capital and labor are so dear on the frontier. A new push to the frontier will unleash a wave of productivity like the settlement of the New World or the American West. But more than a journey of science and engineering, a frontier creates a journey of the soul of humanity. It is high time that we produce more Captain Cooks, not just Captain Kirks.

The essential technical case has rock-solid physics. L5 is a stable, permanently sunny location 60 degrees ahead of the Moon in its orbit. Process heat for manufacturing, cooking, and electricity is plentiful and reliable. All that is needed to conduct modern industry is a big concave mirror that can swivel a little and some mass to work on.

There are a number of technologies that need to be developed. All of them seem, as O’Neill professes, to be well within the bounds of current engineering skills in 1977. I identified a number of technologies and capabilities that need to be developed to make it all work. Any one of them seems pretty easy to develop given sufficient will and $60 billion in 1977 dollars (about $200 billion today):

  1. 3,000,000 tons of Lunar material delivered to L5
  2. Earth launch capability to deliver 10,000 tons of Earth equipment to the Moon and 40,000 tons to L5, or about one 20-ton payload launch a day for six years
  3. A new fleet of 10–20 heavy-lift vehicles
  4. New vacuum manufacturing techniques
  5. New orbital manufacturing techniques
  6. Legal rights of way
  7. Adequate security for the orbital installation
  8. Governance of the 10,000 people in orbit
  9. Installation of a Lunar mass driver
  10. Orbital operations techniques
  11. Credit to pay for the island at L5
  12. Legal and managerial credibility to pull off such an operation
  13. Sociology of a large space colony
  14. Recruitment of the 10,000 people
  15. Training of the 10,000 people
  16. Invention of the new technologies, equipment and processes required
  17. Lunar operations
  18. Lunar prospecting
  19. Lunar construction
  20. Lunar robotics
  21. Lunar smelting
  22. Lunar habitats
  23. Earth-Lunar human transport
  24. Earth-Lunar labor pool
  25. Lunar communications
  26. Lunar navigation
  27. Lunar computer resources
  28. Microwave power transmission
  29. Lunar ore processing
  30. Lunar excavation

Many of the technologies and capabilities described above intuitively seem low risk. If Richard Branson can get 12,000 people to sign up to pay $200,000 for a trip to suborbit and European college kids will accept $6,000 as a salary to spend a year working in an Indian call center, then I think it stands to reason that 10,000 people could easily be recruited to live and work in a space colony. Manufacturing techniques ought not to be too hard to develop since there are new amazing capabilities available to workers with no gravity and plentiful vacuum.

But a few deaths in orbit could easily reverse the migration. A couple of industrial accidents in orbit could attract regulatory scrutiny that might make it as expensive, or even more so, to comply with orbital work safety rules than Earth-bound ones. Even if every one of the 30 technologies or capabilities has an independent 90% chance of being pulled off successfully without undue project delay or cost, the combined chance that all thirty will be successful is only 4%.

A common fallacy in big complex cases is to say that any one argument is suspicious on its own, but when you put it together in a large case that the whole case is more probably true than any one argument to make it. In fact, the chain here is much weaker than its weakest (lowest probability of success) link because if any link in the chain fails, the whole case may be faulty.

Even if every one of the 30 technologies or capabilities has a 90% chance of being pulled off successfully without undue project delay or cost, the combined chance that all thirty will be successful is only 4%.

Another difficult challenge to O’Neill’s thinking is that an L5 habitat is not readily scalable from a small habitat. A small habitat must spin much faster to have Earth gravity so it will cause more discomfort, leading to a correspondingly smaller population that will tolerate it and a lower productivity among those who can. There are high fixed costs associated with a Lunar mass driver that make it difficult to profit from small station construction. Worse, sociology and engineering numbers for a tiny habitat make it difficult for the economics of a small station to work out in the production of space manufactured goods, especially solar satellites.

The lack of scalability means that to achieve O’Neill’s optimistic economics one must jump from nothing to an incredibly large colony. This requires heavily stressing credit facilities, requiring tremendous patience in order to get it built, and ensuring that the many moving parts of the project achieve critical (albeit modest) progress. It is unlikely that private capital will finance such a tremendous undertaking until the risks are retired and the scalability challenge can be addressed.

It is an enticing future that O’Neill offers. There may be a way to salvage it. Looking backward along the critical path of new technologies and capabilities required to achieve success you see a viable Lunar technology, space industry and culture base. The Moon, unlike L5, is much more amenable to scalable technology. Since—as O’Neill himself points out—Lunar solar power stations can be “on a high peak near one of the Lunar poles” (p. 137), L5 does not have much of an advantage in energy reliability and density.

Lunar settlement is also more scalable since habs and capabilities can be built piecemeal.

Technologies 4, 5, 14, 19-27, 29, and 30 are all that is required to establish a Lunar industrial base. That is fewer than half the technologies that need to be developed in order to settle L5. That brings the chance of success up to 23% if the chance for any one of the technologies or capabilities is 90%. Since all of these are required for settling L5, this is not wasted effort. And since only 20, 22-24, and 27 are really required to initiate a Moon outpost, the success of the Moon outpost is much more likely; that lower risk more easily justifies the spending.

Lunar settlement is also more scalable since habs and capabilities can be built piecemeal. There is also a near-term opportunity of science, technology, national pride, and space tourism that pays for Lunar exploration. Since this is already the path we are on with the Vision for Space Exploration, so much the better.

A “phase two” of Lunar development could be hypothesis testing of the viability of an L5 settlement. That would involve technologies 1, 3, 7-9, 13, and 15-18. In all, at the conclusion of phase two, 24 of the 30 identified technologies could be tested and proven before the massive undertaking of building an island in the sky is undertaken.

By retiring the risks, conducting spiral development of technologies required for colonization, and establishing a track record of successful development, operations, law, politics, and economics, steady slow progress can be achieved. The honeymoon may be over for the high frontier, but bringing the case down to earth (or the Moon) will greatly hasten the day that an L5 colony will be funded and succeed.