The Space Reviewin association with SpaceNews

Iland One illustration
The mid-70s estimates for the cost of developing Island One assumed more robust and less expensive space transportation options than seem feasible today. (credit: SSI)

Revisiting Island One

As Professor John Hickman observed in “The Political Economy of Very Large Space Projects”:

While popular science writers typically describe the benefits to be derived from their favorite very large space development project in detail, their treatment of the crucial initial capitalization of such projects is typically sparse or implausible.

Really serious number-crunching is rare, taking a backseat to general pronouncements about what a given piece of technology may enable, and much gushing over the power of entrepreneurs to single-handedly reinvent reality. That they almost never see an actual price tag is, Hickman believes, one reason why so many enthusiasts of space development overestimate the prospects for private efforts by entrepreneurs.

A notable exception to this combination of technological hyping and faith-based accounting is Gerard O’Neill in his work, with his 1976 classic The High Frontier: Human Colonies in Space an excellent example. In Chapter Eight, “The First New World,” he discussed in great detail the financing and logistics of constructing a space colony, the famous “Island One.” This “first world” was to be established at Lagrange Point 5 (L-5), and home to 10,000 people, occupied primarily by the processing of minerals mined on the lunar surface.

This article will examine his consideration of that proposal, seeing how his calculations stack up against the financial and technological realities of space development as they stand today, more than three decades later. The focus will be on one dimension of his analysis: launch costs, which comprise 50–65 percent of the total operation cost in his proposal.

Launching Island One

O’Neill offers a range of estimates of what it would take to get this operation going. At the low end, it would take the transportation of 10,000 tons to L-5, and another 3,000 tons of cargo to get the associated facility on the lunar surface going. The high end is the transport of 80 kilotons to L-5 and another 20 to the surface of the moon. The median estimate runs to 42.5 kilotons at L-5, another 10 kilotons on the moon’s surface.

That they almost never see an actual price tag is, Hickman believes, one reason why so many enthusiasts of space development overestimate the prospects for private efforts by entrepreneurs.

O’Neill envisions the development of a shuttle-derived heavy-lift vehicle (HLV) capable of launching cargo at the cost of $425 (after adjustment to current dollar values) a pound ($937/kg) to Low Earth Orbit. He also assumes that it costs about four times as much as that to get an equivalent payload to L-5 ($1,700/lb, $3,750/kg), and twice as much as that to put cargo on the moon ($3,400/lb, $7,500/kg). The result is that the low estimate runs to $60 billion, the high one to $450 billion.

Of course, O’Neill’s expectations about launch costs (like those of other 1970s-era prophets of space development) proved to be highly optimistic, even given the disagreement about how these are to be calculated. A $10,000 a pound ($22,000 per kilogram) Earth-to-LEO price, almost twenty-five times the estimate O’Neill worked with, is considered the reasonable optimum now.

Going by those numbers, and the relatively greater difficulty of routine trips out to L-5 and the lunar surface that O’Neill suggests, the price tag for getting Island One going would be a staggering $1.5–10 trillion. Even adjusting the costs of getting cargo to the Moon relative to orbital launch downward, in line with the current expectation that it would be priced at only four to six times the rate of cargo delivery to LEO, it would still easily run to $1–6 trillion, an order of magnitude greater than the bill O’Neill anticipated.

There is also the sheer demand such an effort would actually put on the space launch infrastructure to think of. O’Neill posits that the HLVs would need to make three thousand flights over six years in the median estimate. Accordingly, he estimates that ten to twenty shuttle-derived HLVs would be able to do the job.

Of course, cost aside, O’Neill is again working with 1970s-era expectations about shuttle turnaround time, and reliability. Instead of the 25–60 missions a year that it was thought a shuttle orbiter would be able to perform, in practice it has proven to be about three a year. Additionally, the shuttle has demonstrated a two percent failure rate on its sorties, higher by orders of magnitude than originally anticipated.

Assuming these numbers carried over to the higher-performing HLV (which may be optimistic), the median estimate of the lift capacity needed for the projects calls for a fleet not of 10–20 vehicles, but 150 to 200 HLVs, with another 6–10 procured annually to fill in for the losses incurred—almost as many built each year as the entire fleet O’Neill envisioned, devoted solely to this project. (In the case of the high, 100-kiloton requirement, the numbers would come to a work force of 300–400 HLVs, with 13–20 replacements procured annually.)

It stands to reason that the scale of production, the intensity of operations, and the accumulation of experience during this venture would lead to improvements bringing down costs and improving safety and performance as measured by other metrics. Nonetheless, there is no getting around it: where the first set of figures would have been merely prohibitive, the latter set makes the whole endeavor impossible. O’Neill had hoped to realize a profit on Island One in three to six years, but given the bill suggested here, nothing of the kind would have been imaginable—and even a profitable operation would have been forbidding given the capital costs discussed above. Additionally, the accident rate would have been intolerable in a project with the goal of permanently installing ten thousand human beings in a station at L-5.

Making space colonization happen

Some of those reading this may find themselves cursing the Space Shuttle, and NASA, for the millionth time. However, matters would have been no different even if NASA had stuck with the mighty Saturn 5, or if anyone else had used any of the other rocket types actually developed in the three decades since O’Neill conducted his evaluation.

Very clearly, the world’s existing combination of resources and technology will not suffice, which is why no one has undertaken anything close. Changing that will require favorable developments along three tracks. One is the design of a space launcher which is not only cheaper, but given that manned spaceflight is the issue here, also much safer and more reliable than anything built to date.

Very clearly, the world’s existing combination of resources and technology will not suffice, which is why no one has undertaken anything close to Island One.

Another is the downsizing of the plant, both in terms of its actual weight, and the number of people actually required to operate it, a process likely to owe much to both materials science and advances in automation. (See “Diversifying our planetary portfolio”, The Space Review, August 6, 2007) This would, of course, mean compromising one of the main ideas behind Island One, namely the establishment of a substantial population of human beings in space, but this would certainly seem a reasonable one in the early phases of any project aimed at making space development pay its way.

The third is the fostering and sustenance of strong economic growth. Just as the cultivation of a nation’s economic productivity became the key to its military potential in the industrial age, so is it the foundation of the ability of a nation and a species to engage in space development. Given a world GDP three, four, or five times as large as what we have today (all conceivable by 2050, see “Economic growth and space development over the long haul”, The Space Review, September 29, 2008) even a $10-trillion price tag for a space project would not look as intimidating as it does now.

Greater progress in one area, of course, will compensate for slower progress in another. The differing rates of progress in them, however, will do much to decide not just when space colonization happens, but how it happens, by helping to determine who is in a position to foot the bill when this becomes a real option. Those material facts, rather than ideological preferences, are what will ultimately settle the issue over whether private business, national governments, or international collaboration will lead the way.