There is enormous operating and maintenance cost with maintaining the ISS even seperate from transit from and to it, and it would not be responsible to abandon it in place or risk not being able to service it and have the potential for impact and debris polition, especailly in the orbit it occupies. If there is inadequate budget to support the station, it will have to be de-orbited. Currently, the station is only funded through 2020, although the operational lifetime may be able to be extended out to as far as 2028.
As for why the ISS cannot be maintained indefinitely, you need to understand that even in low Earth orbit the space environment is extremely harsh. Between ultraviolet radiation, high energy solar and cosmic radiation, micrometeor and component debris, and even the small but constantly changing tidal forces, the station is under constant attack. Abandoning the station for even a couple of years would likely result in enough damage that it would be extremely costly to repair and refurbish the station for habitation.
With respect to selling or offering the station for salvage, it would be an option to consider, and there is already a nascient industry which is exploring salvage and reuse of satellite components for secondary customers (which saves the mass of having to carry structure and robust components); however, I have yet to see any practical commercial applications for the ISS, or investors who would be prepared to advance the cost with the possibility that if the ISS were rendered inoperable it would have to be written off an de-orbited. It would not really be suitable as a “space hotel” even if such an industry were possible. It is not in a good orbit a fuel depot or for launching off interplanetary missions. It is poorly suited to industrial research or use, and lacks facilities for any kind of asteroid or mineral resource processing. Basically, in order to do anything useful with the station you’d need to haul it into a different orbit and/or boost several additional modules in addition to refurbishment and modification of the existing system, which is likely to be more expensive than just starting from a blank sheet and using more innovative options such as inflatable modules.
With regard to commercial operators such as SpaceX providing faster and cheaper access to space I do not have the same degree of optimism as Sam Stone but I agree with the essential premise that NASA needs to get out of the space transportation business and back into the role of its predecessor (NACA) in serving as a knowledge and funding clearinghouse in order to develop orbital transportation and space exploration technologies which can then be provided or licensed to commercial operators for application. Because of the political angles, NASA is not able to make optimum technical decisions regarding system configuration, and while the SLS is manifestly better than the Constellation system or the STS, it is still a giant pig of a vehicle which by design requires an inordinate amount of ground processing labor with limiited opportunities to achieve cost savings or technology improvements. Ultimately, we no more want NASA to be a primary provider of space launch services than we want a government run airline or or television manufacturer, and the sooner commercial launch providers can become competative the quicker NASA can return to its core competencies; mission engineering, space exploration, and technology interchange.
ISTR a decade or so back, NASA was desperately trying to find i8086 CPUs and compatible components to maintain the shuttles’ computers, because they had not paid attention to Moore, had not anticipated building in an upgrade path, and had been left in the dust by the IT industry. Replacing the shuttles’ very important brains would have been a terrifically expensive undertaking.
The Challenger accident is what got the Shuttle out of the DoD program. DoD didn’t want to tie all their eggs up in 1 basket, and with the Challenger shut down, they needed other options anyway. So that’s when they pulled out as a major payload provider for NASA.
It’s not quite so simple. Rather, due to the bureaucratic processes involved with maintaining and certifying NASA systems, redesigns and system upgrades are necessarily cumbersome and expensive. Far easier to fly another of the same old design than update the design and then certify the design update.
Does it have to be that complicated? Probably not.
The “bureaucratic processes involved with maintaining and certifying NASA systems, redesigns and system upgrades,” are what we call qualification, acceptance, and independent verification and validation (IV&V), and the general rigor with which it is applied is the primary reason why launch vehicles developed under NASA and DoD standards have about half of the failure rate as commercial and foreign launch vehicles. Modern flight software is the single most complex system on a launch vehicle in terms of the breadth of detail requirements and interfaces, and while Norman Augustine (of the Augustine Commission) accurately skewers the costs and schedule slips resulting from ever-increasing complexity flight software, that same complexity is necessary to allow for control of inherently unstable systems and to provide sufficient redundancy for failsafe operation even should components fail. This isn’t like writing code for a website or office productivity application where a failure or reboot means you lose a term paper; an unrecoverable system upset or dead code loop can result in mission failure and catastrophic loss of vehicle.
A major change to avionics or software can have impacts throughout the entire vehicle. Just relocating an inertial guidance sensor can be enough to change response dynamics and cause a previously stable the vehicle to become uncontrollable. A change in processing bit rate can result in the same code being no longer reliable despite having been previously validated. Although there are certainly inefficiencies and excessive redundancy to be found, virtually every piece of paper and check is done because some previous failure to follow sound processes resulted in schedule slip, damage, or flight anomaly. You can get rid of the “complexity” and cost of the documentation and verification processes, but in doing so you also increase risk and the potential for a failure that you cannot even diagnose because of lack of data.
I’m a programmer by training, we studied that in school.
The software for the Space Shuttle, orbiter, ground, the whole package- is the grandest software engineering project in history.
It’s error free. There are no errors in the code, every single bit has been interdependently analyzed and cross checked and reverified.
It’s flawless. It’s legendary. It’s the Hoover Dam of software.
Reporting on the Shuttle always had the tag line,“most complex machine ever built.” But why? Isn’t all this stuff way over engineered? Is it so hard to fly a truck into space that it takes every bit of engineering ingenuity the human race can come up with, resulting in a program we could only afford by getting rid of the most expensive useless military anybody ever saw or heard of? Is our tech just not good enough to design something simple and affordable?
In some very simple terms, and I admit this is NOT the only way to think of the OT, consider this:
Q: What makes a manned rocket fly?
A: Funding. People like Buck Rogers. No bucks, no Buck Rogers.
People don’t want to see manned spacecraft disintegrate on launch or re-entry. People at NASA were aware of damage to the left wing of Columbia that occurred during launch. Executive decisions were made that it wouldn’t be an issue. Not that there was much in the way of in-flight repairs were possible. Similar to the executive decision that SRBs cold soaked way beyond design specification and long established mission rules on the Challenger wouldn’t be an issue. The shuttles were 1970s technology and there’s only so much upgrading and modifications one can do. And the shuttle design was a compromise. NASA didn’t want delta wings but the Air Force insisted. One might note the USAF’s X-37 does NOT have delta wings.
I think it’s a shame that the X-20 Dyna Soar program was cancelled. It could have lead to a better reusable shuttle type craft than what we had.
The NASA budget isn’t all that big - about $17.7 billion for 2013. I know that SOUNDS big, but that’s not all used for building rockets. About 1/3 of that is used for science: building satellites for astronomy and climate science, sending probes to other planets, etc. And they don’t just build stuff, they also pay scientists to operate them and analyze the data. And there is also a lot of basic technology development for space and aeronautics.
That leaves about $8 billion for human spaceflight. But about 1/3 of that is used for operating the ISS, and doing useful things on it (e.g. building scientific experiments to fly on the ISS). Almost $1 billion goes to buying services from commercial space launch companies.
When the Shuttle program was ongoing, it cost almost $4 billion per year. The SLS is expected to cost a total of $30 billion to develop. There just isn’t enough money to do both unless Congress gives NASA a few billion extra dollars a year, which they haven’t.
The $30 billion cost for SLS is unnecessarily high for reasons Stranger outlined throughout this thread. But it’s still less than what it cost Boeing to develop the 787 Dreamliner (reportedly $32 billion). (Incidentally, the annual revenue of Boeing is about 5 times larger than NASA’s budget.)
By the way, I’m pretty sure there are many things around that are more complex than the Shuttle. An Intel CPU in a modern laptop has over 1 billion transistors; that’s probably larger the number of parts in the Shuttle. A modern airliner or aircraft carrier probably has more parts than the Shuttle.
What’s unique about spacecraft is that they have very thin margin for performance, and produced in low numbers, and must work correctly the first time. You can’t take a rocket on a low-altitude test flight, land it and then send it to orbit. (SpaceX, Blue Origin, etc are trying to do just that, but they aren’t in use yet.) The first test flight IS the real thing. You can’t just make everything 10 time stronger than it needs to be, and add a few more engines and tanks for redundancy; it would be too heavy to reach orbit. Also, designing a new rocket and building, say, 3 of them perfectly, is extremely difficult. Things like cars and commercial electronics are reliable because the manufacturer has optimized the assembly line over time until it produces a perfect product every time. (It takes years to build a new car factory and get the assembly line running.) But NASA can’t justify building that many rockets, so they need to ensure reliability the hard way - i.e. by reviewing the design and manufacturing method over and over, and extensively inspecting and testing everything they build, every step of the way.
OK, since we now rely upon the 1960’s design Russian Soyuz spaceship, why don’t we just reactivate the old Apollo program-it would probably cost less that developing a new platform.
Space exploration (manned) is a dead end, until we develop nuclear engine rockets.
I probably seemed a bit more dismissive of the process than I meant to be. I am well aware of everything you said. I work on astronaut hardware.
Okay, here’s the challenge: You have to build a truck, from the ground up, that can haul something the size of a luxury bus, and pull it faster than the speed of sound. You have to build this truck so that it is perfectly safe and that you can prove that it is perfectly safe before you are allowed to let anyone drive it. You cannot take a test drive at slow speeds or with it empty, you have to put a bus in it and drive it full speed on the first go. If your truck gets a flat tire, one random nail on the highway, one speedbump you don’t anticipate, your truck goes kablooey and people drag you out to the public square and draw and quarter you. So you need to think of every possible way your truck could have a problem, and build in a solution that prevents that catastrophic failure. You can’t prescreen the highway and clear all the nails and potholes, so you have to design your truck where a pothole doesn’t send you off the road - at Mach 2.
I’m underselling the challenge.
It is just not possible to pull Apollo out of the closet. It’s not a matter of pulling out a few hundred drawings and remaking steel and aluminum parts. Many of the processes were developed by the contractors doing the work. Many components are obsolete. Finding the computer chips and such just isn’t going to happen. “Well, just replace them with new chips” you say. Like Stranger points out, you still have to go through the entire revalidation program. It’s not just the piping and the tanks and such you have to build, you have to build the fixtures and handling apparatus to allow you to build the tanks and piping. Those fixtures, jigs, and whatnot are not around any more. There is very little cost savings to be had “resurrecting Apollo”. You’re basically building a new vehicle from scratch. You do have the advantage of the lessons learned through Apollo and technologies derived from Apollo (like rocket engine design), but you’re still building a new vehicle.
^^^ plus, Apollo wasn’t much of a spacecraft. Best, most innovative, coolest vehicle that ever carried people to the Moon, yes.
But still the VW Beetle of spacecraft. Functional, but not comfortable for long journeys.
As much as I’d like to have a “new” classic Bug, I wouldn’t pay new-car prices for one.
