Human Settlement of Space: A Strategy for Making It Real,
Including New Options for Energy From Space
Dr. Paul J. Werbos*
1. Defining the Target: Sustainability
Forty years after the Apollo program, we now recognize that the dream which inspired us was not the dream of funding a few short-term visits to the moon, leaving behind only footprints in the sand, at a cost of millions of dollars per footprint. If the space program continues to cost a million dollars per footprint, and if the taxpayer on earth pays the entire bill, we haven’t done what we set out to do. But what did we set out to do? How can we translate our broader vision of humans settling space into a more concrete, operational target that we can use to guide concrete engineering and management efforts?
President Bush’s vision for space exploration rightly stresses sustainability as a key goal in space. Sustainability basically got into the plan as a result of political considerations. Using 1960’s technology, a voyage to Mars would cost many times more than the entire NASA budget, and Congress would not support that; therefore, we know we need to develop new technology, to reduce the cost of space exploration, so as to make it sustainable as a larger-scale activity within a fixed real budget.
This vision was a good start – but we need to take it further. Human presence in space will not be truly sustainable until it is capable of independent economic growth, without needing any net subsidy at all from government, and without being limited by the current rate of growth on earth itself. There are three main requirements which need to be met:
1. Money earned by space (revenue) must grow enough to be able to pay for the entire activity.
2. Activities in space must be large enough and diverse enough that they lead to “multiplier effects,” where people will invest ever more money to support new activities in space paid for by supplying existing activities and people in space.
3. The underlying technology for productive activities in space must be efficient enough that we can “close the loop” economically.
I would propose that the number one goal for human activities in space should be to maximize the probability that we eventually meet all three requirements. Once the goal is accepted, everything else we do should be part of a rational, adaptive strategy to meet the goal – to maximize the probability. The later sections of this chapter will address the question: how do we actually do that? But I need to say more about the goal itself before moving on.
First, the three requirements here are really just an English-language version of a more precise mathematical criterion. The economist Eugene Rostow wrote a famous book decades ago on “take-off economies.” He noticed how some poor nations seem to go from bad to worse; others become “banana republics,” earning money in proportion to the current market for bananas in rich nations, but unable to grow any faster or to catch up; and yet others reach a “takeoff point.” The real goal here is for the human economy beyond the earth to reach the economic take-off point. In more mathematical terms – there is a kind of matrix XS, the portion of the input-output matrix for activities in space which can be economically supplied from space, augmented by considering humans in space as an activity; we want this matrix to have a real eigenvector with eigenvalue greater than one. Because XS is actually a matrix of derivatives, it depends both on existing infrastructure and activity levels in space, and on the characteristics of the technology. In practical terms, this means that we want the infrastructure in space, the revenue-generating activities and the underlying technologies to reach the point where space is no longer just a banana republic cum subsidy farm.
is very hard to assess the total probability that humans will someday meet all
three requirements, across all future history. Thus as a practical matter, we
may approximate this goal by trying to minimize
the expected delay time between now and the time when all three
requirements are met. This is a reasonably good approximation, on the whole.
The longer we wait, the greater the chance that world events will make it
harder to keep going, if we do not reap the economic benefits that space could
give us in the meantime. There are critical technologies we have inherited in
Some advocates would argue that this approximation is not so good after all. They would say: ”The probability of someday meeting these requirements may actually be greater if we focus all our efforts for now on the goal of humans walking on Mars. Footprints on Mars would energize the public, and provide a growth in budgets which makes it easier to reach your requirements.” In effect – a gigantic, expensive public relations effort. But footprints on the moon, under President Nixon, were followed soon after by the largest percentage cut ever in NASA’s budget. Still, there is a legitimate argument here for accelerating the Terrestrial Planet Finder (TPF) mission, drawing on the new minimum-cost technology initially developed by Prof. David Hyland (of the TPF advisory board) under a small grant from NSF. TPF, if fully funded and focused on using advanced technology, will probably be able to see decisive signs of large-scale plant life on earth-like planets in other solar systems. I would guess that this discovery would galvanize public interest in learning more and in strengthening earth’s technology far more than any rerun of Apollo would do (even on Mars). There is room in NASA’s budget to accelerate a few key activities like TPF, even if the bulk of the budget is redirected to the goal I propose here – to minimize the expected delay time between now and the time when the three requirements are met. There is also room for a few lower-cost strategic efforts to encourage the science and technology that may be needed, someday, to maximize our chance of someday reaching beyond our solar system.
greatest convulsions in
Finally, the greatest risk to NASA today comes from paying too much attention to the goal of minimizing the time and cost between today and the time when we are certain that American footprints will appear again in the dust on the moon. We need instead to minimize the cost and time between now and our average best guess of the time when our enabling technologies, discoveries and infrastructure truly allow profit-making entities aimed at real pubic market to “take over” – to operate on a large enough scale, with diverse enough activity, that they truly possess self-sustaining growth in service to humanity.
Notice the big difference between a goal which demands certainty and a goal which aims at the average expected time delay! There is a huge difference between pursuing a difficult longer-term goal, where success cannot be guaranteed, and pursuing a more certain near-term goal. Many prefer the goal of “new footprints in the sand” simply because it can be guaranteed, technologically, by using technologies proven to work at NASA many years ago. There is an analogy here to the large programs on lead-acid batteries at DOE years ago, at the time of long gas lines; everyone knew that lead-acid batteries would never be good enough for real market-worthy electric cars, but they were funded anyway, because they were a “low risk” technologically. We need to remember that there are two kinds of risk – technology risk (when you might not meet your technical goal) and market risk (when the product may not have market value). Re-inventing Apollo would have minimum technology risk, but without a market benefit in sight, the risk of total bankruptcy is very serious. We need to adopt a more balanced attitude to risk, like that of the entrepreneur, who realizes that he must make a tradeoff between technology risk and market risk. He must be bold enough to aim directly at a real market profit, and he must accept the fact that some degree of risk is unavoidable. In the end, the risk that matters is the risk that we never get to our true target. Advanced technology is a way of reducing the risk that really matters. We need to do it right or not do it at all.
2. Energy from Space: New Options and a Critical Need on Earth
The number one requirement for sustainability in space is a huge new revenue stream, aimed at markets on earth, but requiring diverse new activities in space. The most serious candidates right now, in my view, are energy from space (“Space Solar Power,” SSP), space tourism, space manufacturing and materials from the asteroids. In order to implement the strategy of section 1, our civilian space program should be centered about the goal of bringing one or more of these four candidates all the way to commercial fruition. Our efforts must be highly adaptive, because our knowledge about these four markets and the required technologies will change a lot from year to year, if we do our jobs right. If we do our jobs right, we will develop many new technology options and push hard to improve our knowledge by addressing the most difficult challenges and uncertainties surrounding them.
The four big new markets are not alternatives. Down at the level of nuts and bolts, efforts in the four different areas can all help each other. But in this section I will write entirely about SSP, for four reasons: (1) thanks to new progress, SSP shows the most promise of meeting our requirements as soon as possible; (2) this is an area where I personally have special knowledge, as a result of co-managing the NASA-NSF-EPRI joint effort in SSP funded in 2002; (3) SSP can play a vital role in helping us achieve a sustainable global energy/environment system on earth – a crucial high-risk challenge which is a matter of life or death in its own right; and (4) new energy sources motivated by the needs of earth can also play a crucial role as an enabling technology in achieving sustainability in space. In section 3, I will discuss how SSP fits into a larger concrete strategy for sustainable space settlement.
Years ago, I was far more skeptical about SSP as a flagship market for space. I still have great respect for those people who remain skeptical today, based on the options discussed back in the 1970’s. In the 1970’s, NASA and the Department of Energy (DOE) were both funded by Congress to evaluate SSP. NASA funded two major studies to develop “reference designs” for SSP which, they estimated, could produce 24-hour (“base load”) electricity at 5.5 cents per kilowatt hour (kwh). That is roughly competitive with coal and nuclear power, whose total cost now runs at about 4-8 cents per kwh today.
The DOE study was more pessimistic. The DOE study was led by Fred Koomanoff, of the Germantown Office of DOE. Fred brought in assistance from several places – including me, since I was then the person in DOE Headquarters (the Office of Energy Information Validation, located in the Energy Information Administration) responsible for evaluating all aspects of long-term energy projections and their assumptions. Our report was far more skeptical than the NASA report. It questioned the cost estimates, and it also pointed to major unresolved uncertainties about whether the technology would work as advertised. Unfortunately, Ralph Nader and others misrepresented this report as an attack on the idea of exploring SSP at all; in fact, that was not the intent. A rational and honest assessment of uncertainties and realities is an essential first step to making a technology real – if there is any hope of making it real at all. Koomanoff and I both tried to be clear on that point, but Nader’s voice was of greater interest to the media.
In the 1990’s, thanks to Congressman Rohrabacher and John Mankins, research on SSP was revived at NASA. Mankins’ “SERT” program achieved a number of very solid accomplishments, essential to the new hopes we see today. To begin with, Mankins did verify that some of the problems we were worried about would have been show-stoppers, in the original reference system designs. But then he moved on to develop new solid, verified technologies that would overcome those problems. He obtained detailed reviews and recommendations from the National Academy of Sciences, whose report is still an important document on SSP. He also developed serious cost estimates for the best new designs which resulted from this effort. Unfortunately, by about 2002, the best serious cost estimates were just under 20 cents per kwh – good enough for some niche markets, but not good enough to compete with coal or nuclear fission for the big baseload power markets on earth.
two important new developments arose. First, we led a new workshop, joint with
John Mankins, on how we might use computational intelligence, learning and robotics
to try to reduce the cost of SSP. That workshop report, still available on the
web site of Prof. George Bekey at the
That study asked key decision-makers and policy experts all over the world what is the number one contribution that science and technology could make to improve the human condition in general; the number one answer was a non-fossil non-fission source of affordable baseload electricity on a scale large enough to meet all the needs of earth. (See www.stateofthefuture.org .) SSP is one of the very few serious candidates to meet that need. And then, in 2001, my current employer – the Electrical and Communication Systems Division of NSF – was approached by one of the leaders of the microwave communication community (Prof. Michael Steer of North Carolina Sate University), whom we funded to do a quick reassessment of the microwave power beaming issues associated with SSP.
sources gave us very strong encouragement to go ahead and try to start up a
funding activity in this area. The leaders in microwave communication told us
that the challenges of power beaming and of its supposed hazards were all
basically matters of engineering and verification – but little if any objective
risk is present in the underlying technology. We can do it. (We later funded a
study by Frank Little of Texas A&M University which provides the best
current information on this aspect.) And so, in 2002, James Mink and I from NSF
approached John Mankins of NASA to propose a 50-50 new joint venture, to be
routed through the NSF machinery for funding universities and small businesses,
with an open call for proposals distributed all across the
This solicitation was published as NSF publication “number” NSF 02-098, still available on the web, with references to prior work and to the funding priorities as we saw them in 2002. We announced that we only had $3 million to hand out – but we still received a very diverse and impressive collection of 98 proposals. After very intensive peer review, demanding that we only consider truly original and unique work with a major potential impact on the economics of SSP, our panels recommended funding of about 50 proposals, amounting to $21 million. There is a major unmet opportunity here, unmet to this day. We deeply regret the very valuable work proposed that we could not fund as yet – but even so, we learned a great deal from this exercise. John Mankins and I jointly ran the exercise (though I had the administrative duty of coordinating panels and allocating proposals and so on). We both had very intensive discussions with the various groups, as their work progressed, and we both learned a great deal about the realities. With help from NASA’s Jet Propulsion Laboratory (JPL), we jointly wrote a “technology profile” on SSP included in the 2003 technology book of the interagency Climate Change Technology Program – the only relatively official summary of the overall (interim) outcome of the SSP effort.
So where do we stand today on actually making this work at an affordable price?
We now have what I would call “four new design concepts,” all much more promising than the 17-cent SERT options. These are:
1. A SERT-plus design, a concept from John Mankins, similar to some
of the previous designs, but embodying new “solar sandwich” ideas from Neville Marzwell of the Jet Propulsion Laboratory (JPL) and other innovations.
2. The “spinal cord laser” design which Richard Fork (
3. A hybrid laser/D-D design which I first proposed in the State of the Future 2003 (CD ROM appendix).
4. The not-so-new or not-yet-new-and-ready concepts for using nonterrestrial materials as proposed by Gerard K. O’Neill, David Criswell and others.
None of the four now offer such a credible price tag as the SERT designs, simply because we have not had the support yet to flesh them out, evaluate the uncertainties and develop sharp cost estimates. But in at least three of the four cases, we have excellent reason to expect that costs will be much less than 17 cents when we get there. There are important risks – but when we consider all four options together, I believe that the objective risk of not being able to beat coal or fission on cost is far less here than with any other technology I have looked at large enough to meet all the world’s energy needs.
For an updated version of the global energy story, see the companion web page here.
(This does not mean that we should not fund those other “team B” options, but we should not count on them. And earth cannot afford that we neglect our best option for baseload alternative power.)
On the whole, I would guess that SERT-plus would end up around ten cents per kwh, if we meet certain assumptions built into the SERT cost estimates. The one really worrisome assumption, in my view, is the assumption that the cost of access to low earth orbit (LEO) will be cut to $200 per pound. At $2,000 per pound or more (the best we can realistically hope for with expendable rockets), that gets raised to $1.00 per kwh – strictly a niche market. The key concept here is the use of lightweight mirrors (like those proven out by Entech) to focus light onto “sandwich cells,” which produce electricity beamed to earth by microwave.
Spinal Cord Laser
In the “spinal cord laser” (which Richard Fork initially called the “backbone laser”), there is almost no flow of electricity at all in space. Lightweight mirrors concentrate light onto disks of semiconductor material, which are excited by sunlight in the same way that photovoltaics are excited. These disks also exhibit similar efficiencies and weight, but emit energy in the form of coherent light which can be beamed directly down to small receiving stations on earth. The same laser can easily punch through clouds on cloudy days. Concentrator cells like Marzwell’s could be used to extract electricity on the ground. Because of the similarity to photovoltaic designs, and because we avoid the weight needed for electric power distribution (about half the weight of the SERT designs), it is reasonable to expect costs in the 10-20 cent range. One beauty of this design is that it can be tested and used at a much lower minimum power level than the SERT designs.
My hybrid concept is riskier than the other two, but still has an excellent chance of working out, and promises much lower costs than the other two. The idea is to start with a different light-to-light laser, analogous to the spinal cord laser, but harder to design. We need a laser which can emit energy in the high-speed megajoule pulses needed to ignite nuclear fusion in the deuterium-deuterium (D-D) pellets designed by John Perkins of Lawrence Livermore Laboratories. One way of looking at this is that Perkins’ pellets augment the power production by a factor of 100 or more, beyond what came out of the laser. Thus if the required lasers end up weighing three times as much as the spinal cord laser, for the same output of light, we still end up increasing power output per pound in orbit by a factor of (100/3), or 33. That implies a cost of well under 1 cent per kwh, for electric power available at a central point ready to be beamed down to earth.
From another viewpoint – the hybrid design has major advantages over nuclear fusion on earth. There are two mainstream versions of nuclear fusion for use on earth – fusion in magnetic bottles, and fusion based on lasers hitting pellets. Both versions are expected to cost more than today’s nuclear fission, because the energy comes out as energetic particles which turn into heat and radioactivity very quickly. They require the same complex heat systems as today’s fission and coal, and the raise the same kinds of issues in handling radioactive materials. Building the lasers on earth imposes many, many additional costs. But by doing this on space, and using D-D pellets, we end up with
80-90% of the fusion energy coming out as electric currents – protons moving through the vacuum of space. Instead of a heat chamber, we only need lightweight magnets, to transform the electricity into a form we can beam down as microwave energy to earth. Several credible leading laser laboratories say that they know how to design such a laser, relatively quickly, using new concepts such as photonic bandgap materials, fiber lasers and variations of Raman backscattering. But certainly, the characteristics of the new laser are the main objective uncertainty here, and we need to nail this down as soon as possible (as constructively as possible). To make this option certifiably real, we have two urgent needs: (1) to work on the access to space issues; and (2) to create a small competitive funding venue to develop, simulate and evaluate competing laser designs from universities, small business and laboratories all over the earth, if possible. Even at $2,000 per pound, however, this option looks as if it will come in well under 10 cents per kwh. If the laser costs meet our present best guess, then the costs of delivered power on earth could be competitive with existing baseload costs – i.e., about 5 cents per kwh—if the cost of power beaming itself is 4 cents per kwh (as estimated today for off-the-shelf designs) or lower (as many believe could be achieved with further research on power beaming as such.).
Finally, the fourth option – SSP based on nonterrestrial materials – is not so well explored as these others. Because of the hope of deep cost reductions, the need for nonterrestrial materials in the long-term anyway, and because of other technology benefits, this option cries out for much more effective strategic analysis and pursuit. The cost estimates available today are far less reliable than the crude guesses here for the other three – but we need to work to change this. The costs will still be proportional to the earth-to-LEO cost, because the main cost is still the cost of lifting a lot of material into space. Also, a great deal of testing will be needed to clarify the options and make it possible to estimate costs. This option was discussed further in our year 2000 workshop.
In summary, to approach the normal costs of baseload electric power in the big markets on earth today, we will either need to develop an advanced form of SSP (options 3 or 4) or develop $200/pound access to space, or both. If we develop both, energy from space could become a true bonanza for earth and space both.
mainly addresses the goal of developing space – but SSP is equally essential to
the strategic picture on earth. There is not sufficient space to do justice to
that complex issue here – but I can give an overview and some citations. In
2003, at a workshop in
At present, hybrid cars (up to 60mpg “city”) and plug-in hybrids (using half as much gasoline, but also using overnight electricity) are the only force actively moving us towards independence from OPEC oil – but they aggravate the already-serious problem of where we get the baseload electricity from. Earth-based solar farms are the only renewable energy source which – like SSP – could easily meet all the world’s energy needs, but they only work at daytime; storage to provide energy at night adds a lot to the cost, and the going market cost today is already about 12 cents per kwh in the daytime (though we have ways to reduce that). From a purely technical viewpoint, I see no reason why SSP could not provide truly massive amounts of electricity as soon as 10 years from the start of a truly massive, focused effort – soon enough to prevent the kind of disaster we are now on track to encounter by about 2025. For more information, see the State of the Future, or the companion web page here.
3. What Would an Optimal Space Policy Look Like?
According to the goals of section one, the “optimal” space policy – the best possible space policy – is the policy which gets us to fulfilling the three requirements in the shortest possible time, subject to budget constraints. To get to such a policy, we would need to consider several key principles.
should not leave it all to NASA. The goal of space settlement is not just a
goal for NASA. Conversely, humanity has other important goals – like energy
sustainability and security – which should also concern us all. There are
exciting opportunities to “kill two birds with one stone” by expanding
collaborations between different government agencies and others. For example,
true commercial enterprises can help provide guidance to NASA on how to do a better
job in supporting the four key markets discussed in section 2. The military and
intelligence agencies have great needs, essential technology and funds for
improved access to space and related goals. The National Science Foundation –
in partnership with NASA, as in NSF 02-098 – can effectively solicit ideas from
a wide spectrum of universities, disciplines and small businesses, and provide
more novelty and feedback than NASA could manage on its own, as well as an
agile funding mechanism. Lawrence Livermore and other components of DOE would
have to play an important role in implementing my concept of energy from space.
International partnerships could expand all this in many directions; for
example, robotic and telerobotic technology from
Second, while these programs need to be led by champions who truly care about long-term success, they should not be led by advocates who believe in tilting their message in a way that creates a less than accurate image, in order to elicit support. When programs are not led by sincere champions, the funds have a way of disappearing into activities which support the leader’s constituents, other goals, or nicely packaged but insufficient efforts. On the other hand, when programs are led by advocates, there is an equally long history of ultimate failure. When “team players” filter information and do not face up to key difficulties early on, it is common that these difficulties surface too late for people to explore alternative pathways and save the program. Advocates, publicists and honest worriers should always be “in the loop,” because of their creativity, but they should not be in charge.
Third – no matter which big market for space we focus on, the biggest single barrier today is still the cost of getting to earth orbit. If present trends are continued, I worry that all versions of space development will be doomed to failure in the end. The private sector can help, but it cannot work miracles. There are very tough technical barriers that require the very best enabling technologies, to drive the costs down enough. Sheer scaling laws say we need a big vehicle to get the best dollars per pound – and we cannot afford anything less than the best. Furthermore, we need to remember that the market for access to space has two very different segments – the segment which can get by with payloads of ten tons or less, and the segment which requires heavy lift.
For the ten-ton market – NASA has known for a long time that operations costs are the dominant cost in getting to space, and that aircraft-style operation will be necessary. (For example, see the 1960’s reports from Mueller at Goddard, which inspired the idea of a space shuttle – a great idea, ruined by political confusions and wrong metrics from the bean-counters.) This then requires horizontal-takeoff reusable rockets, at least until more exotic alternatives become available. In fact, at NSF I funded some very aggressive research on some advanced alternatives, through Ray Chase of ANSER and Miles of Princeton University. A startling outcome was that really advanced alternatives can become real only if we first resurrect and enhance some endangered legacy technologies. Ideally, this means that we really need to develop a major program to insert those new technologies into a new program for a 1.5-million-pound horizontal-takeoff rocketplane. The most crucial endangered technologies are “hot structure” technologies which would allow such a big rocketplane to return safely to the earth without using the weird and problematic tiles or ablative structures or slush hydrogen that other approaches would rely on. Ray Chase has developed a detailed plan for how to do this, using off-the-shelf technology all the way, with a project planning chart that goes out just 4 years – though we would be better off starting from a detailed one-year tradeoff/design-refinement study for $900,000, as proposed by Chase. Some have claimed that space elevators would be better than any kind of spaceplane – but before we can afford to build space elevators, we will need affordable transportation to carry the required material into space!
In a truly ideal world, competitive launch services contracts from NASA, DOD and the intelligence agencies would give the big aerospace companies the resources they need to go to Wall Street and finance the new vehicle on a private sector basis, after smaller start-up contracts refurbish the hot structures technology and support the $900,000 study. In addition, launch service contracts for heavy lift capacity at 60 tons payload would be used to allow companies like Space Island Group (Gene Meyers’ company) to build a shuttle-C vehicle enhanced by bringing the expended fuel tanks into orbit. This would cost much less than the presently planned heavy lift vehicles, and help NASA redirect funds to the development of the rocketplane and other key needs. “Meyer’s ark” is a truly exciting concept, even though it (like Criswell’s) requires a bit of fine-tuning.
of a permanent human presence on the moon certainly plays a critical role –
along with the asteroids – in opening up a source of lower-cost materials which
will be essential, sooner or later, to the full economic development of space.
But development of the moon must be part of the larger development of space
described in section 1, in order to contribute the most it can to the long-term
goal. Stronger infrastructure will be needed both in earth orbit and in low
lunar orbit (LLO) to make this possible. Again, we need to do it right as soon
as possible – and that means doing it with efficient infrastructure, and not
“saving time” and wasting money by doing things the wrong way. If 10-ton payloads end up costing $200/pound
(and lower, with time), and if heavy lift costs ten times as much or more, it
will be essential to economize on how we use that heavy lift capability.
Reynerson of Boeing has described concepts of a totally reusable rocket (RLAV) to go to and from the moon; he
reports that an RLAV can be quite efficient even weighing only 60 tons, if it
is launched dry and unmanned. (Fuel and people can be carried aboard the
cheaper rocketplane.) Kistler’s new lunar rocket company has proposed similar
ideas. Long-term fuel depots in LLO are well within today’s best technology.
Ion rockets have been proven out in deep space by
* The views herein are those of the author and do not reflect the official views of his employer; however, it does constitute work by a government employee performed on government time.