Skip to main content

The Mega-Module Path to Nowhere (Or: How to Eliminate Human Space Flight With an HLV)

The recent Ad Astra Onlinearticle by John Strickland, "The MegaModule Path to Space Exploration (Or: How to Use an HLV)" is a sadreflection of old-school thinking. His argument for heavy lift is the sameapproach adopted in the 1960's, which made Project Apollo a costly dead end.

Strickland calls for NASA to"to go for the largest available booster which can be created at areasonable cost." On the surface, that seems reasonable--who could beagainst reasonable cost? But Strickland never bothers to quantify the cost ofthose super-boosters and mega-modules or what "reasonable" cost is.As Robert Heinlein said, if you can't say it in mathematics, it isn't science;it's opinion.

The Saturn V cost about $2billion per flight, in 2005 dollars. NASA estimates it will cost $10 billion to developits big Shuttle-derived Heavy Lifter. After that, we can expect each launch tobe about the same cost as the Saturn V. This expense is due to high capitalcosts (expensive infrastructure to support such super-boosters) and high laborcosts (due to the huge workforce needed to launch them). Furthermore, NASA'sShuttle-derived HLV will use explosive solid rocket motors that require carefulhandling, which may drive costs up even more. (The late Dr. Wernher Von Braunemphatically rejected the use of solids on the Saturn rockets.)

Common wisdom in theaerospace community is that larger rockets will be cheaper than smallerrockets. Yet, common wisdom has failed to reduce the cost of human spaceflight--why?

"Common wisdom" isbased on the payload fraction--a larger rocket can carry more payload for eachpound of liftoff weight. So, a larger rocket should cost less per pound ofpayload. That argument's fine, as far as it goes, but it ignores other factors.Most of the liftoff weight is propellant, and rocket propellants typically costpennies per pound. For current systems, propellant cost is less than 1% oflaunch costs, so optimizing for mass fraction alone doesn't produce low cost.

We need to look at capitaland labor costs, which account for 99% of launch costs. With expendablerockets, there's a certain amount of capital that's expended with each launch--therocket itself--but there's also a huge amount of fixed capital, in the form oflaunch pads, vehicle assembly buildings, factories, etc. The size and cost ofthe launch and manufacturing facilities depends on the size of the rocket.

Basic business theory saysthat when you have a large capital investment, you should use it as often aspossible to get the maximum productivity out of it. If you're going to investin a launch pad, for example, you ought to use it as often as you can. Yet,Strickland calls for just the opposite--"avoiding multiple launches."That means the pads and other facilities will be underutilized. In principle, alaunch pad can be used at least twice a week. (The late Dr. Max Hunter launchedtwo Thors from the same pad within 48 hours.) Yet, NASA's plan calls for it tolaunch just three heavy lifters per year.

Assume two expendablerockets, Hoss and Little Joe. Hoss can launch 100-tons. Little Joe can launchonly 10 tons. Hoss requires $6 billion worth of launch pads, roadways, hangars,etc. Little Joe requires only $600 million worth of launch facilities. If youlaunch Hoss three times a year for 10 years, the amortized cost of those launchfacilities will be $6,000,000,000/30 = $200,000,000 per flight, or $2,000,000per ton of payload. To do the same job with Little Joe requires at least 10times as many flights. Let's call it 12 times, to allow for Hoss's moreefficient payload fraction. So, Little Joe must be launched 36 times a year.So, the amortized cost of launch facilities over 10 years will be$600,000,000/360 = about $1,670,000 per flight or $167,000 per ton of payload.

Labor costs show similaradvantages for Little Joe compared to Hoss. Hoss might require 5 times as manyworkers to assembly, checkout, and prepare for launch. Hoss is launched onlythree times a year, but you have to keep the team together between flights. So,each worker must be paid 4 months wages for each flight. Little Joe, on theother hand, is launched about once every 10 days, so workers only get 10 dayswages for each flight. Annual labor costs for Hoss are 5 times what they arefor Little Joe, to deliver the same payload.

There's more. The Little Joecrews are launching 36 rockets a year. The Hoss crews are only launching 3rockets a year. So, the Little Joe crews are a lot more practiced in theirjobs. The more often you do something, the more efficient you become.Economists call this "the learning effect." If you perform a tasktwice as often, efficiency generally improves 10-20%. The Little Joe crews arepracticing their jobs 12 times as often, so their efficiency should be at least30% better.

These factors lead to asingle, inescapable conclusion. The most important factor effecting launchcosts is not the size of the rocket. It's the flight rate.

The advocates of great bigrockets point to 747s as evidence for economies of scale favoring largevehicles. But smaller jets like the 737 carry far more passengers, withvirtually the same economy per seat-mile. Airlines know that economy scalesmost strongly with flight rate, not aircraft size. 747s operate only on majorroutes where there's enough demand to fill a 747 on a daily basis. When you considerthese factors, it's little wonder that a Planetary Society group headed by MikeGriffin wrote that "present launch costs of $9-11K per kg will continuefor the next 40 years" when they recommended the super-booster/mega-modulearchitecture. Their assessment is a classic example of a self-fulfillingprophecy.

There is not enough demandto fly a heavy-lift vehicle on a daily basis, or a weekly basis, or even amonthly basis. Nothing in NASA's plans, or the plans of private industry, evencomes close. Someday, when we are launching thousands of tons of payload everyday, heavy lift vehicles will make sense. Until then, vehicles should be sizedappropriately for current demand.

But what about right now--whatoptions does NASA have?

Space Adventures and itsRussian partners are offering circumlunar flights for an estimated $100 millionper flight. For the cost of developing a Shuttle-derived super-booster, NASAcould buy 100 circumlunar flights from Space Adventures. Or, if you assume alunar landing is five times as hard as a circumlunar trip, they might buy 20Moon landings for the same price. If Russian hardware is not politicallyacceptable, Boeing and Lockheed could develop systems to use their existingDelta and Atlas rockets, using the multiple launch techniques once proposed forthe Lunar Gemini program.

Even a strident advocate ofheavy lift like Dr. Robert Zubrin has acknowledged that low-cost missions couldbe done with current vehicles. In "The Case for Mars" he writes that"even if Energia or other Russian launch vehicles (such as Proton...) arenot available or allowed, the mission still doesn't have to cost that much.Current launch costs using existing US boosters such as Titan's, Atlas's, orDelta's is about $10,000 per kilogram. At these rates... launching the entire300 (metric) tonnes would cost $3 billion. Add that to the $1 billion forhardware development, and once again we find a total cost of less than $5billion."

With orbital rendezvous anddocking (proven long ago in the Gemini program and many times since in Apollo,Skylab, Mir, and ISS), we can build space structures far larger than the piddlylittle 100-ton "mega-modules" Strickland writes about. With existingexpendable launch vehicles, NASA could send humans to the Moon and Mars and doit *sooner* than the current timetables call for.

At the same time, NASA canleave the door open for the future. Consider, for a moment, a small reusablelaunch vehicle, like the Space Van proposed by Len Cormier of Third MillenniumInternational. The Space Van could carry a crew of two plus 8 passengers or aton of cargo into orbit, for less than $500,000 per flight.

Just one Space Van, flyingfive times a week, 50 weeks a year could launch 250 tons a year. That's moremass than two of Strickland's "mega-modules," at a cost of $125million instead of $4 billion. A fleet of 12 Space Vans could launch 1500 tonsa year -- the equivalent of 15 Skylabs or 15 Strickland mega-modules -- plus3200 human beings.

Cormier believes the SpaceVan could be developed for $200 million. That seems amazingly low by expendablelaunch vehicle standards, but there's good reason to believe he's in the rightball park. Remember, this is a small vehicle, and there's some evidence thatreusable vehicles are cheaper to develop than ELVs. A large part of developmentis testing; reusable vehicles are cheaper to fly, so they should also becheaper to flight test. In the 1960's, General Dynamics did a study of the X-15and the Atlas A, reusable and expendable vehicles of similar performance. Theyfound that a piloted, reusable vehicle was more complex but also 40% cheaper todevelop.

SpaceShipOne providesanother example. The flight test program included 66 flights of the first stage(White Knight) and 17 flights of the upper stage (SpaceShipOne). If these hadbeen expendable stages, Scaled Composites would have had to build a total of 83stages to complete the same test program. If each expendable stage cost $5million, the total program would have cost nearly half a billion dollars.Because both vehicles were reusable, Scaled was able to complete the flighttest program for only $25 million.

Of course, afirst-generation RLV like Space Van might not pan out. Many new vehicles arenot quite right on the first attempt--the de Havilland Comet airliner, forexample. But, even if it takes two or three generations for developers to getit right, the cost would still be a fraction of the $10 billion needed todevelop NASA's super booster.

NASA could encourage thedevelopment of such vehicles by designing its lunar and Mars missions around amodular, open architecture, with many small components that can be easily movedbetween vehicles. In the short term, this would allow NASA to negotiate betterprices from competing launch vendors. In the long term, it would allow NASA totake advantage of new vehicles that don't exist and may not even be envisionedtoday.

Or, NASA can mortgage itsfuture by tying itself to "mega-modules" that can be launched only bygiant super boosters, which only the government can afford to develop.

Ultimately, the choice comesdown to what kind of future we want to have. Mr. Strickland promotes heavylifters as a way to avoid multiple launches and develop orbital fuel depotswith "no permanent human crew needed." I prefer to reduce the cost ofspace access so that we can have human crews. I want to see a future where wehave a lot *more* launches and a lot *more* people in space.

NASA's proposed lunararchitecture would allow three moonshots a year, with four astronauts on eachflight. This is just 12 flight opportunities per year, down drastically fromthe Shuttle program. It's no secret that NASA's contemplating astronautlayoffs. What kind of message does that send to America's young people, whohave been told to study hard, excel in math and science, so they can grow up tobecome astronauts?

Small reusable vehicles canrevolution access to space the way microcomputers revolutionized access tocomputing. NASA can help enable a vibrant future where thousands of peopletravel into space every year -- or it can cling to the super-boosters and mega-modulesof the "mainframe" era, in which case the NASA manned space programwill continue its slow downward spiral. As a classic science fiction movieasked, "All the universe, or nothing--which shall it be?"

Edward Wright is presidentof X-Rocket LLC, a company specializing in spaceflight services, training, andeducation.

NOTE: The views of this article are the author's and do not reflect the policies of the National Space Society.

Visit Astra Online for more news, views and scientific inquiry from the National Space Society.

Join our Space Forums to keep talking space on the latest missions, night sky and more! And if you have a news tip, correction or comment, let us know at: