The Once & Future Moon
Smithsonian Air & Space
A constant refrain in policy discussions is that the space program is too expensive. This verity is at the core of the Augustine report, which asserts that Project Constellation (NASA’s Program of Record to implement the Vision for Space Exploration) is too expensive and requires at least an additional $3 billion per year to implement. The “New Space” community is aggressively campaigning for commercial launch services, pronouncing any NASA-designed/NASA-built space system politically “unsustainable” due to the agency’s inability to receive adequate funding from Congress over the lifetime of a flight program.
The civilian space agency’s 2010 budget is about $19 billion. That’s a nice chunk of change but compared to what the federal government spends elsewhere, it’s not a head-turning sum. For comparison, the Department of Energy’s nominal funding currently stands at $26 billion, to which Congress recently added an additional “one-time” $35 billion for the next two years, averaging out to about $43 billion per year – more than double what NASA will receive.
Other federal agencies and departments are funded at even higher levels. The Department of Education’s FY 2010 budget is $68 billion, of which $56 billion is “discretionary” (meaning non-entitlement). This handsome outlay is for an agency whose principal mission is to influence and monitor public schools; in fact, educating tomorrow’s citizens is the responsibility of state and local authorities and is covered by those taxes, which make up approximately half of the states’ budgets. So in federal spending terms, NASA does not consume as much as most people believe nor is it even “expensive” when stacked up against the cost of other agencies. And when managed even moderately well, we get something for our investment in NASA.
What drives space costs? Advanced technology alone is not the cause. An iPod contains more advanced systems than any single piece of electronic equipment carried 40 years ago on an Apollo spacecraft and is available for a couple hundred dollars. An array of electronic boxes in our homes offer 500 channels of high-definition video in surround-sound stereo – a stunning visual and aural sensory assault – for only a couple thousand dollars. The family automobile contains sensory and monitoring computers, protective airbags, fuel injectors, catalytic converters, automatic parallel parking systems, GPS, satellite radio, DVD players, and a slew of other innovations only dreamed of 20 years ago. These vehicles are affordable enough that most families have more than one in their driveway and buy new models on a regular basis.
Building and launching space vehicles is expensive but the reasons why might surprise you. It’s not the equipment or even the infrastructure that drives up costs. It may well cost hundreds of millions to billions of dollars to build a launch pad and its associated facilities, but once constructed and maintained, it is used essentially forever (some facilities at the Cape have launched rockets for over 50 years). The propellant used to hurl payloads into space makes up only about one percent of the cost of launch. Rockets are built of shaped aluminum (along with a few other more exotic metals) and those pieces make up an additional 10% or so of the total cost.
Spaceflight costs remain high because it requires complex machines, with millions of parts working together in a precise order and in perfect coordination to put a payload into space. To assure that these events transpire as planned, we pay a large number of highly skilled technicians, engineers and scientists to design, build and operate space systems. These high demand people don’t work for minimum wage. It requires almost 10,000 people to operate the U.S. Space Shuttle launch vehicle system. Everyone has a critical job, from program design to inspecting and replacing thermal tiles on the orbiter airframe, to stacking and configuring the vehicle for launch –- everything and everyone necessary for the construction, operation and flight of the vehicle to and from orbit. It is a specialized machine that is custom-made and individually operated. Moreover, each individual piece of hardware has a paperwork trail so that part failures can be tracked back in time and space. Documenting these part histories and pedigrees requires many hours, all billed to some charge code.
Despite our best efforts, rockets are finicky and unpredictable. Sometimes, satellites don’t wind up in the proper orbit or fail to operate correctly. Customers who paid for the satellite need some kind of indemnification against these possibilities. Insurance rates are based on a careful assessment and determination of risk. For a launch system with a long history of reliable performance, premiums are relatively low (but still substantial). New rockets and new companies face higher insurance costs and it may take many years to establish a track record of enough resiliency and consistency so as to significantly lower insurance rates. These costs must be folded into the cost per kilogram to LEO.
Which brings us to the inescapable fact that a major obstacle to routine affordable spaceflight –what makes other high technology efforts affordable – is the lack of mass production and automation in the fabrication of space systems. If we could mass produce rockets and automatically assemble and check them out for launch, launch costs would drop dramatically. Commercial items are inexpensive because development costs are amortized over very large sales volumes. For space systems, development costs are very large and not easily hidden by amortization.
A way to make spaceflight cheap is to remove much of the highly paid human talent from the end-to-end processing stream. One possibility is to automate most of the process of rocket fabrication, assembly, checkout and launch. A wholly new approach to our launch service infrastructure and model of operations would be required and to my knowledge, no company or government entity is working on such an approach. Even SpaceX uses a skilled cadre of people to custom build, launch and operate their vehicles. The claimed goal of SpaceX is a factor of ten reduction in cost and increase in reliability. I hope they reach it. But even if they do, space travel is still a costly enterprise; reduction of cost from the current $10,000 per kilogram (Atlas 5) to $5400 per kg (the quoted current cost for a Falcon 9 launch, which is not yet operational) is progress, but is not the canonical “hundreds of dollars per pound” breakthrough sought by space fans everywhere.
Another way to lower costs is to do what others in high-technology fields have done: outsource the work overseas. The Indian PSLV rocket can put 3700 kg into low Earth orbit and costs about $20 million (at least, that’s what informed sources claim it cost to launch TECSAR, an Israeli radar imaging satellite). This price works out to be about $5400 per kg, exactly the same as the projected cost for the Falcon 9 – and the PSLV already has a proven track record. The Russian Proton rocket puts about 20,000 kg into orbit for an estimated $115 million, about $5700 per kg. Even the supposedly “costly” European/French Ariane V puts 18,000 kg into space at a cost of $120 million, or about $6600 per kg.
The cost of all these competing systems seems to be approaching a single value – $5000 per kilogram is achievable for launch within the existing engineering state-of-the-art. Is such a cost “cheap enough?” Commercial launch costs have hovered between $5000 and $10,000 per kilogram (constant dollars) for the last thirty years. This “expensive” price structure has given rise to a thriving commercial space industry, especially in global communications. And despite the hype about orders-of-magnitude decreases in the cost of launch, these numbers likely will persist for the indefinite future. SpaceX has no access to special physics, ULA cannot repeal the law of gravity and XCOR cannot change the rocket equation.
Space launch costs what it costs. Spaceflight is expensive because we employ and pay thousands of highly skilled and trained people to build and operate space systems. Despite decades of planning and talking, these costs have not decreased significantly and the dream of cheap space launch remains a chimera. We frequently get marginal improvements in the dollars per kilogram number, but never of the order-of-magnitude variety. The problem of small volume/high cost cannot be solved by factors of two or three decreases in launch cost. We need a new operational approach that severs the Gordian Knot problem of the cost of space launch.
The use of off-planet resources of materials and energy is that approach. Despite new launch vehicles, new companies and supposed new approaches, we have only marginal improvements in the cost numbers for launch. It is time for the new and fundamentally different approach of developing the Moon’s natural resources to build a space faring infrastructure that will create new capabilities and give us lasting value for our money.
Paul D. Spudis is a Senior Staff Scientist at the Lunar and Planetary Institute in Houston, Texas.
