Two papers blended into one for the House Space Committee. Unfortunatly they didn't take my advice. Oh well.








Jerry E. Pournelle, Ph.D.

Chairman, Citizens Advisory Council on National Space Policy

Testimony to Subcommittee on Space


March 16, 1995




Side Benefits 5



SSX 11


Shuttle Promises 12





This document has been revised to include answers to some questions asked at the March 16 hearings; in particular, why Shuttle made all the promises now made by SSTO; why is the SSTO program different from the Shuttle program?

A formal reply to questions sent after the session is attached.



NASA was a Cold War creation. Now that the Seventy Years War is over, why is NASA needed?

The potential of space is very great, but at present there is high risk and little immediate payoff. The major market for space services is government; while private markets will develop, they haven’t yet, and probably won’t until access costs are lower. There isn’t a great deal of private incentive to bring those access costs down because the markets have not been developed. Lowering access costs and developing space markets requires a long term commitment. The return from that will be high, but it’s in the future.

The discounted value of a dollar in 15 years is effectively zero.

What will you give me if I promise to pay you $1000 in 15 to 20 years?

It isn’t reasonable for private companies to do long term investments.

Bell Labs where the transistor was invented was special and doesn’t exist any more. There is nothing else like it. We have no institutions charged with long term technology development. This puts the United States at a distinct disadvantage, because not all nations have our limitations on futurist development.

We all know that sometime in the next century, space will be very important to the international economy. It’s not rational for any private company to do space development research. Someone must. That leaves government.

Adam Smith held that enterprises in which the risk is high, the return to all is great, and the benefit to any one investor is problematical, are proper subjects for government attention. This is a good example: the return is high, but more to the next generation than this one.

The legitimate mission of NASA is to do long term technology developments in aeronautical and space sciences; and to make the results available to American companies for exploitation.

In other words, the purpose of NASA is to look ahead of the profit window and identify promising aerospace technologies for future development.

The way to demonstrate new technological discoveries and identify the technologies needing further research and development is through X projects.



X Projects incorporate state of the art technology into a project which focuses on a technological objective. Their mission is to stretch the envelopes of technological knowledge. They may achieve other results.

The primary mission of an X Project is technological.

An X space ship would not have a mission or cargo requirement beyond what is needed to achieve the technological objectives.


X Projects are not prototypes.

We don’t expect to build a lot of them.

X Project craft are expendable.

We don’t want to crash an X ship, but it can happen, and we cannot be so concerned about the possibility that we don’t get on with the project. Typically three X ships are planned. One is often expended. The second is then flown given the knowledge learned by the loss of the first. Sometimes additional models are built to reflect discoveries made by flying the first ones.

X Projects are short term. Ships are planned, built, and flown. Then the project ends.

X Projects build upon each other. They may overlap.

The purpose of X projects is to demonstrate new technologies, learn about new capabilities, study operations with the new technologies, and discover new areas that need work.

X Projects may not lead directly to prototypes. It is a mistake to insist that they do so.

The best example is the X-1, which was intended to demonstrate the possibility of supersonic flight. Flying faster than sound was an elusive and costly goal. The X-1 was built with the single purpose of demonstrating that it could be done. There was no cargo other than the pilot.

The X-1 did what it was intended to do. It also gave new insights into the problems of flying in transonic regimes.

It was followed by the X-3 Stiletto, a plane that could not have been built when the X-1 was planned. The X-3 did lead to operational aircraft, including the F-104 Starfighter.

The X projects were greatly successful. They were effectively ended in the late 1960’s. The X programs were not canceled in the name of economy. Knowledge gained through the X programs helped U.S. aerospace firms to dominate the world industry. In the 1970’s US high technology, particularly aircraft, were the largest single cash export of the nation. They were very important in making up the deficits in our balance of payments.

The X projects were canceled because of arms control. The arms control strategy held that rapid development of military technology was undesirable since it led to ‘new rounds in the arms race.’

Whatever the wisdom of the arms control strategy in the Cold War, there is no arms race now. There is instead a new international competition which resembles the Cold War in that it is ‘silent and apparently peaceful, but could well be decisive’. It is a new Technological War, and the United States has a great deal to gain from achieving not merely a leading, but a dominant position in all areas of technology. In particular, we can achieve a commanding lead in aerospace technology. The way to achieve that is through properly organizing NASA into the leadership agency in long term aerospace R&;D, and reviving the X programs.



There are side benefits to X programs.

One benefit that may not be obvious is career continuity. Because commercial firms must look to short term profits, it is often necessary to downsize in unprofitable years, and hire extensively when business is good. The effect on high technology careers can be devastating. Our recent history looks like a bad parody of manpower resource allocation.

It takes a while to learn how to do aerospace engineering and project management. There is also value in having been part of a project team. X projects provide a small but significant number of technological people with continuity of employment doing something worth doing. Talent lost to the aerospace industry is generally lost forever.

X projects are a source of hands-on experience. They are not jobs programs; the X projects are themselves valuable. They are also fairly short term, and generally carried out in places like Edwards Air Force Base, China Lake, White Sands, and other areas unlikely to attract people who want to build bureaucratic empires. X programs generate a talent pool of experienced people likely to be hired away from government service into private industry at need, but available during periods of low economic growth.

Note again we are not advocating a jobs program; merely continuity of employment for a small number of key people doing important work. Think of a good X program office as the government’s Skunk Works. Its very existence is a major plus factor in international competition.




The key to space development is to get lots of people into space. Once they can get there, they’ll develop markets. Bring resources, energy, and ingenuity together and the result has always been wealth. Space has resources and energy. What it doesn’t have is entrepreneurs with access to the space environment.

There’s no reason space cannot be a source of revenue. This is important when government runs at a deficit. If we’re going to borrow money from our grandchildren, it’s both reasonable and ethical to spend some of it on genuine investments. Space can be, but has not been, a place for genuine investment: that is, a potential revenue source, not a perpetual sink.

The way to get lots of people into space is to bring the cost of space access down by a factor of about one thousand. It can and should be brought down to costs comparable to long distance airline operations. In the next section we’ll look at how that might be feasible. For the moment, consider what would happen.

There isn’t much market for space services because the cost of getting there is so high. Even so, there are proposals for 800 + communications satellites, to be launched and maintained at the present ridiculously high costs, with the full expectation of profit. Weather satellites are worth the costs of putting them up. There are national security payloads that are worth present costs.

Not much else is. Hubble Space Telescope is returning some wonderful pictures, and there are numerous science experiments that give us very good data, but it’s questionable whether the data are worth the costs. On the other hand, if space missions cost millions instead of hundreds of millions to billions, everything changes. The spacecraft themselves would be cheaper, because there would be no need to design them to last unattended and forever. Spacecraft tend to cost about the same as the operations costs of getting them up. This is sensible, but in our rapidly changing technology environment, the electronics in spacecraft are generally obsolete long before the spacecraft has ceased operations.

Future space markets are totally dependent on access costs. They are also unpredictable. Take an example from airplane days. Suppose in 1920 the Congress had tried to form an intelligent estimate of the economic potential of airline travel; in particular the number of tickets that might be sold. One probable route would be New York City to Los Angeles, California. They might look at the number of people taking that trip by train. They’d then factor in the ease of travel by air as opposed to trains, and try to guess at a number. If they felt very bold they might decide that as many as 500 a week would take the trip. Then they could be extravagant and multiply that by two, to get 1,000 a week. They might even go mad and estimate ten thousand a week.

They’d never come close to the actual numbers on any reasonable or even sane set of assumptions, and even if they went mad and guessed the right numbers, no one would believe them, and they’d still not have a handle on the second order effects: the industries that are only made possible by rapid travel capabilities.

It’s the same way with space. When space access gets down to the price of first class airline travel, it’s nearly impossible to estimate the volume of business.

The simplest business is tourist travel. I have asked travel agents to consider an ocean trip to, say, Grenada, with the high point of the trip being four to eight hours in space—a round the world trip with a vengeance. How many would buy tickets depends on the ticket price, but even at $50,000 a ticket, the estimates of the number of tickets that could be sold are surprisingly high.

The story is the same with science and industry. Get the cost of space access down and the volume of traffic goes up sharply. A few years ago the cellular phone was science fiction. Then it was a status symbol. Now we can contemplate every citizen having a personal telephone number that doesn’t change no matter where the temporary or permanent residence. In another generation that won’t be a prediction but a necessity.

Twenty years ago G. Harry Stine described some potential space industries in his book The Third Industrial Revolution. I described others in my A Step Farther Out. None of these marvels came to pass because the cost of access to space remains so high; but given reasonable access costs those industries would develop very rapidly.

The military implications of low cost access should be obvious. "Information Warfare" is the new buzz phrase in the war colleges. Whatever it means and whatever you think of Information Warfare, one thing is obvious: if you are going to control information, you must control access to space, and you’ll need to get there a lot cheaper than we do it now.

I have previously described the system I called THOR. Consider a tungsten rod about twenty feet long and a foot in diameter. Put it in orbit, and have a means to direct its reentry. Give it a terminal guidance system—fins, or use a means to move the center of gravity—and a Global Positioning System receiver. Don’t bother with a warhead. The result will be a missile able to hit any point on earth with an accuracy of under 25 feet and a closing velocity in excess of 12,000 feet per second. The result is energies comparable to tons of TNT buried under the target. Few structures: bridge abutments, fuel dumps, anchored warships including both battleships and carriers, fuel dumps, hardened armor parks, etc., can withstand that.

Observation and communications satellites will be available to theater commanders on the same basis as AWACS flights.

All this is not only possible but inevitable with low cost access to space. The only question is who will get there first to take advantage of the space environment. At the moment the United States has a lead in space access, but other nations can do this analysis as easily as we can, and many have done so. Several are looking at ways to gain low cost access to space.

There is a way to reduce the cost of access.




Airlines operate at a small multiple of fuel costs. This includes both flight operations and amortization of aircraft. It takes about the same fuel energy to put a pound in orbit as it does to fly that pound from Los Angeles to Sydney, Australia. Since rocket engines are as efficient as jet engines, there is no reason why space operations should cost more than two or three times what long distance air travel costs. In particular, you should be able to buy a ticket to orbit for no more than twice what it costs to buy a ticket to Sydney.

Of course airlines don’t push the airplane off the end of the runway into the Coral Sea when the airplane gets to Australia; nor do they rebuild it. They do routine service, refuel it, and fly it back. That is what space operations must be like if we are to make access to space affordable to the American people.

The obvious way to go to space is to fly to space and return; to operate like an airplane. That wasn’t possible with the materials and engines available in the early days of rockets. Nearly everyone is agreed that it is possible now.

We know that SSTO is possible, and we know how to build a single stage to orbit ship. However, we don’t yet know how to design operational SSTO craft. We know we can build SSTO ships, and there are probably several ways that will work. Fundamental designs include: ships with wings; ships without wings; lifting body ships that take off vertically and land horizontally; ships that take off vertically and land vertically; ships that reenter nose first; ships that reenter tail first; and variants of the above. Fuels range from hydrogen to propane. There are several possible engine configurations.

Most—probably all—of these will ‘work’ in the sense that the ship will get to orbit; what isn’t known is how much cargo the ship will carry. Note that I say cargo rather than ‘payload’. Payload is a term that comes from the days when we threw the ship away: there was structure, which was bad, and payload, which was good. In those days ships were designed to maximize performance, and performance was measured by payload to orbit.

That is no longer the best way to design ships. SSTO ship designs should be driven by operations, not performance. Take space construction as an example. We can design ships to deliver one large payload; but suppose we can build the structure at far lower total cost by taking up smaller payloads at vastly lower costs per flight. Even if doing so requires redesign of our primary structure the resulting savings can be in the billions. This is what I mean by operational rather than performance driven design.

At the moment we can’t really estimate operations costs, because we don’t really know what SSTO payloads will be. The rocket equation—rocket science, if you will—says that the amount of useful cargo in a Single Stage to Orbit ship will be a third decimal fraction of the Gross Liftoff Weight (usually abbreviated as GLOW). That is, if I have a ship that weighs 500,000 pounds at takeoff, about 450,000 pounds of that will be fuel and oxidizer. Of the remaining 50,000 pounds, between 30,000 and 50,000 pounds will be ship structure and the fuel required to bring the ship home. What’s left over is cargo delivered to space. That cargo can range between zero—none—and about 20,000 pounds, depending on how light we can make the engines, tankage, and structure of the ship.

However, it’s not really as simple as that, because we aren’t sure that we’ll need the full 450,000 pounds of fuel and oxidizer; it depends on drag. Drag is the term used to describe air resistance as the ship rises. Drag is critical because it affects everything else. The lower the drag, the more quickly the ship will lift. The faster the ship lifts the more quickly it gets to higher altitudes. The higher the ship gets, the more efficiently the fuel burns in the rocket engines, and, since the air is thinner, the lower the drag.

In other words, my assumptions about drag are crucial to determining my predicted cargo.

We do not have accurate predictions of drag. We’re going to have to fly some ships to get the required accuracy.

Similarly, we don’t know some crucial facts about stresses which determine the structural mass of the ship. We need to fly ships to find out.

There are other unknowns. The bottom line, though, is that we need some more flight data before we can design operational Single Stage to Orbit ships. Until we have real flight data, the cargo—payload weight, if you prefer—is determined by the assumptions you plug into your computer model, and no one set of assumptions is markedly better than another.




We need an SSX-1; an experimental single stage to orbit space ship. That ship should be built to have these characteristics:

Be Savable

It should survive an engine out on takeoff.

Fly Often

It needs to be flown many times, including going to orbit twice in one day.

Fly Soon

Build the SSX with known technology. Its purpose is to identify what we need to know, not to be part of a longer technology development. We need some hard numbers about thrust and drag and control surfaces and command authority and other such technical matters. We need those numbers soon.

Fly Higher and Faster

SSX need not get to orbit, but it should fly high and fast enough to experience reentry conditions; and it should be designed so that if all goes just right it will make orbit. Orbital flight should not be precluded by the design. On the other hand, we shouldn’t be concerned about cargo or ‘payload.’.

Note that an SSX can be incrementally tested. The first flight can be partially fueled and last only a few seconds. After analysis of flight data you can fly again, this time higher and faster. The test series gives new data at each stage. By the time you do a maximum fuel and thrust test you will know a great deal more about the ship’s capabilities.

Make the SSX savable, build it with a high safety factor, then reduce structural weight as flight test data show what are critical stress areas and what are not.

An SSX-1 program would cost about $1 billion and take about four years. While a billion dollars is not trivial, it’s a pretty small investment in what we all know will, some day, be a business comparable to air travel.



Shuttle Promises

The original proposal for Shuttle as a National Space Transportation System used many of the arguments now heard for Single Stage to Orbit. Shuttle didn’t lower launch costs. Why will this be different?

Not Shuttle II

We don’t here propose that NASA build Shuttle II, or any other kind of ‘National Space Transportation System.’ We don’t believe that any single SSTO design can possibly become the national launch system any more than any single airplane design could become ‘the national air transportation system.’

Shuttle was designed to employ about 20,000 people. It met that goal admirably; you can’t fly Shuttle with fewer people. It just can’t be done.

Airlines typically operate with about 110 employees per airplane, and about half of those sell tickets. High technology systems like SR-71 had about 50 people per airplane. Shuttle has nearly 25,000 (including contractors) for 4 orbiters. At an average of $100,000 per employee, we have a fixed overhead of over $2 billion a year. Assume ten flights per year and the cost is $250 million per flight before adding in variable costs like fuel, new engines, and other operations. Since Shuttle needs all those people, there’s no chance of getting the cost per flight lower than, say, $350 million. Actual operations costs appear to be considerably higher than that. If Shuttle had 50,000 pounds of payload per flight then the minimum cost of a pound to orbit is about $7,000. Most estimates are that it’s higher.

Operations Not Performance

Shuttle was designed under the older rocket design philosophy of maximizing performance. In the days of throwaway rockets and disintegrating totem poles this made sense; but it doesn’t any more. We should instead design to minimize operations costs. If this means smaller payloads per flight, then so be it. Our new SSTO vehicles -- there will be more than one kind -- should be designed for operational simplicity and minimum operations costs.

X Systems, Not A National Transport System

We do not here advocate a new National Space Transportation System, or a new National Launch System. We don’t believe there ought to be such things, and if there were, NASA shouldn’t build and operate them. That isn’t NASA’s value to the American people.

Imagine a ‘National Air Transport System’: a single kind of airplane for all aircraft applications from crop dusting to long distance passenger sevice to logistics support of South Polar operations. Clearly that’s absurd. It’s equally absurd to postulate a ‘National Space Transportation System.’

There must be no Shuttle II.

What we advocate is advanced R&;D culminating in X ships; in particular the SSX. SSX will investigate technologies required for operational SSTO ships. We’ve listed the SSX characteristics before, but it’s worth repeating them:


Multi-engine ship that can survive engine out on takeoff.


Get the ship flying within four years. Sooner if possible.


Fly it a lot, through a number of flight regimes. Get it to space twice in one day.


As high and fast as we can with easily obtained technology.

In addition to the SSX-1 program, there should be a parallel effort to develop a new engine. The new engine should be reliable and reusable like the RL-10; throttleable, again like the RL-10; and cheap. The RL-10 is not cheap at present because we buy very few of them, and there’s no incentive to make them cheaper. It wouldn’t be hard to get their costs down: simply commit $40 million to buying RL-10 engines and ask for competititive bids on how many can be supplied for that price. You will certainly get more than 100 for that price.

NASA can then use these engines in X Space Ship programs. It cannot be too often stated that X ships are neither operational ships nor prototypes of operational ships.


Let Industry Choose the Operational Ships


NASA shouldn’t be an operating agency. NASA shouldn’t be flying the ships. Why should NASA dictate ship design? NASA should develop new technologies, and let industry design and operate the ships. Certainly NASA ought to cooperate with industry in selecting what technologies ought to be developed; but that’s not at all the same as doing the actual designs.

We already know how to build Single Stage to Orbit ships. We don’t yet know the best designs for operational SSTO. Let NASA investigate the technological unknowns. Industry will do the rest.






Space station as presently conceived has few potential tenants.

It remains one of our few high technology R&;D programs.

The proper way to think about Space Station is as a learning process: as a series of X projects designed to develop space construction technology.

In particular, Space Station should be designed to make use of on-orbit assembly by assembly crews: not by Ph.D. astronauts, but by the equivalent of deep sea oil riggers. Working on an oil rig platform is hard and exacting work, and deep sea assembly is every bit as technically challenging as capturing a satellite in zero gravity; but we don’t insist that everyone who works on an oil rig have a Ph.D. in Aerospace Science from MIT. Space station should begin the process in which all Americans get a chance at space access.

If Space Station becomes a series of space construction X projects we will learn about station construction; US industries will regain the lead in orbital technology. That is the proper use of Space Station. Stop worrying about its missions and who will make use of it; use it as a means to regain American leadership in space construction experience. Space is an environment we need to know more about. Use Space Station for that.

Simultaneously, we must continue the X programs that lead to low cost access to space; and we have to learn how to supply space installations at reasonable costs with operationally affordable payloads. Seen properly, Space Station is part of our overall strategy to gain a commanding lead in space operations technology.




NASA should stop worrying about Shuttle Replacement. Something will replace Shuttle, but it won’t be a single "National Transportation System" nor should whatever it is be operated by NASA. Many parts of NASA have been obsessed to the point of paralysis by what they perceive as the problem of Shuttle II. They should be told to stop thinking about it and get on with the primary mission of NASA.

NASA should be forbidden to plan Shuttle Replacement Systems. That is the only way to end NASA’s obsession with Shuttle II.

NASA should get out of the flight operations business. This means Shuttle I. It also means Shuttle II. As a general proposition, by the time we know enough about space transportation systems to design an operational system, we know enough to turn that problem over to the engines of free enterprise. In the past NASA has stifled free enterprise by trying to keep a monopoly on space operations—to own the "National Transportation System." That must never happen again.

NASA should remember that it is also the primary long term R&;D agency for aeronautical systems. NASA and before it NACA had a long and admirable record of facilitating aeronautical developments through provision of wind tunnels, ranges, and other test and analytical facilities. They should renew those activities.

The NACA model should also be the model for space development.

NASA can, by focusing on R&;D in space technology, be instrumental in aiding US industry to provide the American people low cost access to space; and to giving US industry, and thus the United States, a commanding lead in what Speaker Gingrich has called "The Greatest Frontier."





birdline.gif (1428 bytes)