The U.S. Space Transportation System
(The Space Shuttle)
The Space Shuttle Program is coming to an end. This represents the end of an era for humanity, for the United States, and for NASA. For thirty years, manned access to space for the U.S. (and cargo access for heavy modules) has been the purview of this program. In the past decade, U.S. astronauts have also gone into space aboard Russian Soyuz spacecraft, mostly to travel to and from the International Space Station. For those of us 45 years of age and under, though, the Space Shuttle really represents the first and only form of American space travel that we can recall.
Why is it ending? There are numerous answers. Certainly the main components, the Space Shuttle Orbiters, are not technically near the end of their designed service life. Although they are in some cases over thirty years old, they have flown many fewer flights than were originally planned, and are perhaps the most coddled, inspected and maintained pieces of machinery ever built. They were originally intended to fly 100 missions each over a 10-year lifespan; instead, they've flown many fewer (~24-39 for the survivors) but over a much longer time. There are many reasons cited for allowing the program to end, but generally they can be reduced to two general categories: cost and safety.
The Space Shuttle was originally projected to cost much, much less per launch than the U.S. is paying now. The costing analysis was done originally for a purely equatorial-launch, civilian spacecraft flying at a high rate, mostly to construct and then service a permanent space station. It didn't turn out that way, though. Even before it was built, the Space Shuttle designs were modified to accomodate the needs of one of its biggest planned customers - the U.S. Department of Defense. That organization wanted a spacecraft which could lift, rendezvous with and service large satellites in circumpolar orbits - spy satellites, basically. This meant a space launch system which had to generate over a thousand miles per hour of additional delta-V at launch, as it wouldn't be able to use the earth's rotational velocity when on a polar mission. It meant a system which therefore had a lower payload and higher fuel fraction, which couldn't be changed - meaning the equatorial missions, too, had reduced perfomance. It was a space launch system which was to maintain two separate launch facilities to do this - Vandenberg AFB on the West Coast for classified DoD polar missions, and Kennedy Space Center on the East Coast for NASA missions into equatorial orbits. All of this meant more money than was planned (although Vandenberg ended up never launching a shuttle due to the STS-51L DoD halt).
Then, to make matters worse, it turned out that the U.S. simply couldn't (or wasn't willing) to fly as many missions as the system had originally been planned for. Note that while this in many cases didn't up the total amount of money spent, as a great deal of that money was spent maintaining physical plant and labor force, in the case of consumables (external tanks, SRB refurbishment, etc. to name only a couple) the price per instance did climb dramatically, as entire industries were maintained to produce many fewer products than planned.
After the STS-51L accident and the loss of Space Shuttle Challenger with all hands, the DoD decided to pull all of its payloads from the Space Shuttle program and rely solely on their own launch vehicles. That deprived the program of nearly half of its planned 'revenue' - further driving up the costs to NASA and the civilian space program.
Whereas the original launch costs were projected to cost as little as $100 million each at the program's outset, it swiftly became clear that this number was low. But how much of the cost was variable (per launch) costs, and how much was fixed (base) costs? NASA studies generally quote a GAO report from 1993 which looked at this exact question. It found that in 1994 dollars, the 'zero base cost' for the Shuttle Program (i.e. what it would cost to maintain the system for a year while flying no missions) was approximately $1.875 billion/year. If the program made five launches per year (which was historically accurate, allowing for the post-STS-51L shutdown) then the total cost of the program ended up being around $2.5 billion. That means that the five flights would cost an average of $125 million each (these are in 1994 dollars). That's already high, even considering the difference between 1980 (planning) and 1994 (actual) dollars - but it's not terrible. Moving forward to 2008, however, we find that in 2008 dollars, the Shuttle was consuming $2.5 billion per year in fixed costs, and variable costs of $825 million - or $165 million per flight. (Source - NASA using GAO numbers).
Yes, it's expensive. Space travel always has been. While the Apollo Program may not have cost near those same numbers, remember that Apollo was designed and flown in the 1960s. Also remember that NASA's fixed infrastructure as an organization isn't reflected in those numbers (HR, procurement, accounting, finance, IT, security, etc. etc.) and would exist no matter what program was used for Human Space Flight. Those numbers also do not include the cost of the payloads, which are assumed to be constant regardless of which launcher or program flies them (i.e. the satellites, laboratories, experiments etc. carried up).
So. It turned out expensive. But it didn't turn out that much more expensive on a per year basis than originally thought - perhaps $1 billion or so more per year. Flights turned out more expensive, but not much more expensive than originally planned - $165 million/flight (2008 dollars) vs. around $90-100 million/flight (1980 dollars). If you include fixed NASA costs, however, given the low numbers of launches, the cost per launch ends up somewhere around $1.3 billion per launch in 2005-2006 dollars - a number that opponents of the program love to cite.
Safety has turned into a much larger concern. We've lost two Space Shuttles - OV-099 Challenger on mission STS-51L at launch January 28, 1986; and OV-102 Columbia on mission STS-107 during re-entry February 1, 2003. Both were lost with all hands. Is the space shuttle unsafe? Well, yes. All space travel is inherently unsafe. Average mission projections for loss of the vehicle due to an unpredictable micrometeoroid strike are around 1 in 275. We've flown 133 flights as of the time of this writeup, with 2 additional flights to go before program end. Assuming those return safely, we will have lost 2 orbiters over 135 flights - a failure rate of approximately 1.5 percent. This is higher than was allowed for, and it's a sobering number to contemplate if you're watching astronauts board an orbiter.
It's not that out of the norm, however. The Soviet/Russian Soyuz manned flight program, at present the only alternative for manned flight to the ISS, has lost at least four flights; Soyuz-1 crashed on re-entry, killing the sole cosmonaut on board, Vladimir Komarov. Soyuz-11 lost life support on re-entry, killing all three cosmonauts on board. Soyuz-18a malfunctioned during ascent, nearly killing the crew upon landing several thousand miles downrange. Soyuz-T-10-1 (the 48th flight) exploded before launch, with the crew surviving due to the Crew Escape System.
More astronauts were lost in the Shuttle accidents due to the Shuttle's crew capacity of seven or eight rather than three. But in any case, the Shuttle has turned out to be less safe than projected, and certainly much less safe than hoped. The real question, though, is should that prevent us from flying it? The astronauts who fly it certainly don't think so, or they wouldn't have volunteered to ride it time and again, many of them on multiple missions. As NASA personnel like to say, "Space is hard." The Shuttle makes it harder still - as it is asked to do a much more difficult job than simpler systems like Soyuz, it is much more complex. By most estimates, the Space Shuttle (especially when sitting on the pad as the full Space Transportation System) is the most complex machine ever built by man, with millions (not thousands) of component systems and parts. A frightening number of those systems are what NASA calls Criticality-One or Criticality-One-R, meaning that they are single points of failure or have only one backup between them and the loss of an orbiter or crew in the event of failure.
Complexity and Technology
Why is the Shuttle so complicated? Well, it has a complicated job. The advantage of using disposable space capsules is that you only need to bring the minimum number of systems along with you to do the job for each phase. You throw away the big, complicated booster and its complex engines during ascent. You don't have to bring nearly as many control systems, because all you're doing is falling on re-entry. The Shuttle, on the other hand, has several sets of systems within the Orbiter:
Unlike the more efficient capsule, the Shuttle carries all of these systems through all phases of the mission. Even though you don't need a big heavy landing gear in orbit, you have to carry it there with you in order to have it available when you come home - so every pound of descent systems has to be carried up, meaning each is a pound less of payload you can carry. Once you're in orbit, the SSMEs and associated systems turn into dead weight, not even used during landing - and again, that's all weight that you 'paid for.'
It's natural to ask why, then, the Shuttle was designed this way. There are a number of answers. First of all, there are a lot of systems (life support and on-orbit systems) which would need to be sent up every mission - and they're expensive to build. The Shuttle, unlike a capsule, is not just a transportation system. Even without habitat modules in the cargo bay, the Shuttle can be used for mission durations of up to two weeks during which astronauts can work within the shuttle as well as use the airlocks and heavy EMUs carried aloft to perform useful work outside the ship itself without forcing all the crew into suits during EVA.
It's perhaps ironic that the Shuttle's lift capacity, intended to help it build and service a space station, was for the first half of its lifespan not used for that purpose. There was no space station planned or built, and so the Shuttle instead acted more like an orbital RV - it took humans up into space and gave them a place to live for a couple of weeks while they did Science and engineering experiments. Was this the most efficient way to do that science? Probably not. But it's what was available.
Once the International Space Station got underway, the Shuttle was suddenly doing what it had been designed for. It could carry heavy components of the station aloft, and did; it could serve as a work outpost for the assembly task spacewalks throughout the process, which it did; it could serve as a crew living quarters for the short-duration crew during its missions. It could finally fulfill its design purpose - it was a delivery truck, and it performed admirably, carrying huge quantities (in space flight terms) of cargo and consumables both up to the station and back down to earth.
But what now?
Part of the problem with maintaining the Shuttle is, again, its age. It is hideously complex, and its systems were designed (most of them) in the late 1960s and 1970s. As a result, it is becoming increasingly difficult (and exponentially expensive) to maintain spare parts for it. Not just suppliers, but whole industries have vanished (try to find a CRT monitor rated for space, see how well you do). When you can't feasibly change the Orbiters to use newer tech without complex, time-consuming and very expensive upgrade programs (the Glass Cockpit upgrades are the prime example) then you reach a point where maintaining the system becomes first difficult, then tortuous, and finally impossible. Add this to the fact that these large complex machines were only expected to last 10 or so years, and you can imagine the increasing cost and complexity of just doing regular maintenance on the beasts.
No; it's time for the Space Shuttle to retire. After all, its predecessor the Apollo program only flew for six years! Thirty years is an unprecedented run, especially for something this complicated.
What next? This is the question that haunts those of us who are firm advocates of space exploration and space travel. At present, the United States and NASA seem to be focusing on using the government space agency to do three general tasks. One; maintain and continue unmanned science and engineering flights (which don't have anything to do with the Shuttle and never did, so no big issue here). Two; foster commercial industry solutions to sending humans into low earth orbit using whatever systems are cheapest, safest and most efficient. In other words, letting industry revisit the decades-old technology of space capsules with modern tech and materials to see if it can be done smaller/faster/cheaper. I have no doubt it can be, but again, this is merely doing something we've been doing since the 1960s, smaller/faster/cheaper. Three; work on designing interplanetary transfer missions and craft. These would be (likely) assembled in orbit to send astronauts to the Moon and other bodies in our solar system. As NASA's chief scientist Waleed Abdalati said during a talk recently, "It's time for us to do things we haven't done before."
While that may be true, it reveals something about the Shuttle. The Shuttle, everyone seems eager to say, was a massive overreach. It was a complex mission with complicated requirements, leading to a goldplated and hugely inefficient engineering solution which still wasn't as safe as desired. Yep, that's all true. But think about it - the Space Shuttle worked. It sent a vehicle into orbit which could be landed like an aircraft and reused. Did it do it efficiently? No. Did it do it as safely as we'd like? No. But it did it.
Maybe that means that in a few years, just as the capsules are getting their second wave of technology now, the manned shuttle will be something that corporations and governments decide is a more efficient way of doing business if modern technology can be applied. There are signs of hope: the U.S. Air Force is flying a spacecraft called the X-37B which is best described as a 'miniature automated space shuttle.' It's a small craft, and is entirely unmanned. Nor is it large enough to carry a working lifesystem. On the other hand, it is a spaceplane - it re-enters and returns to Earth using aerosurfaces, and lands horizontally on a runway. Even more impressive, it lands itself - something which the Buran, the Soviet Space Shuttle program, may have done once decades ago, but something the Space Shuttle can't do today for various reasons.
Although lots of people don't believe them, the Air Force claims the X-37B is just an experimental testbed. I hope they're telling the truth. Because if they are, what they're testing are the very technologies we (NASA or industry) would need to build another, better, cheaper, safer Space Shuttle. They're testing lighter and more reliable thermal protection systems. The X-37B doesn't have hydraulic systems for its control surfaces - rather, it relies on electric actuators (motors) which are lighter, don't require heavy and bulky fluidics, and are better able to withstand the 'cold soak' of space. It has computers and avionics capable of remaining in orbit for up to 270 days, changing orbits multiple times, and then autonomously computing and executing a landing (perfect the one time it has happened) at a designated spot on earth, softly. Imagine what we could do if we took the various technologies there and started over with the Space Shuttle!
Will we? Who knows. One thing is certain - the Space Shuttle was ahead of its time. It was built to be a delivery truck moving between the surface and LEO, and therefore assumed that there would be places to go in LEO that mattered. There finally is, now - the ISS. The ISS could be used as a base to assemble modules brought up by heavy-lift rocket or by shuttle systems for use as interplanetary or extra-orbital vehicles. Additional space stations, smaller and more efficient ones, might be built using what we've learned from building the ISS - and as we learned there, having a vehicle which can transport and house humans while they work is a prerequisite for space construction.
There is a sadness at the core of this, though. For thirty years, humanity and the United States have been flying spaceships. Working, honest-to-god spaceships with pilots and crew. Now, we're about to lose that, to go back to being cargo rather than crew and go back to spam-in-a-can rather than the glory of a spaceship. Sure, it'll be more efficient and cheaper and maybe safer. But to the ten-year-olds who will be our future, will it be as exciting? Will it make them see themselves in spacesuits and helmets, exploring new places? Or will it make them see space as just an expensive place to get to and to work like any other? If the latter, we lose.
And we can't afford that.