Space Shuttle - pros and cons, a summary

[nothke]

(picture: Empty status board in VAB, after the last Shuttle mission, credit: Penny4NASA)

Since some of the people are crying and some celebrating after the retirement of the Space Shuttle, I wanted to clarify was the Shuttle any useful and why was it made at all if it wasn't. We can see from facts that it was too risky and too expensive, but that is a consiquence, not a con, the con is the REASON why it's risky or expensive. I have searched for answers on the internet and this is a short summary of what I have read.

Pros

Reusability - Ok, this one is obvious, reusability of the Orbiter and SRBs freed the System of continuous production (which is the case of for eg. Soyuz). So that cut the costs.

Flexibility - The shuttle was designed according to requirements set by the military. The crucial factor in the size and shape of the Shuttle Orbiter was the requirement that it be able to accommodate the largest planned commercial and military satellites, and have over 1,000 mile cross-range recovery range to meet the requirement for classified USAF missions for a once-around abort from a launch to a polar orbit. The ability that differentiates so much from other spacecraft is the ability to retrieve (bulk) cargo from orbit, for ground repairs or for retrieving experiments. It has been used for deploying, repairing and retrieving satellites and the Hubble, performing experiments in space and docking to and building of a space station and crew rotation.

An engineering feat - It was a symbol of the entire nation and a symbol of it's technological supremacy. It was really sexy I must say.. (I'm not American)

Experience - This is more like a side effect, but in my opinion, it is big (right?). Running the same thing for 30 years made NASA know it better with each flight. Only Soyuz could be given the title of the most "experienced" system

Cons

Too complicated - Apparently, this is the biggest con of the Space Shuttle. It just had too many sub-systems and so too many things could fail, and too many things to check and prepare for next mission. For comparison, the heat shield of a typical pod manned rocket was composed of one single ablative heat shield, while the Shuttle had 7 different TPS systems, each guarding the surface according to how much heat it takes during reentry. Also note that 2 of the highest resisting materials (the glass foam tiles and reinforced carbon-carbon leading edges and nose cone) were very fragile. The complicated system added greatly to risk and cost.

Lack of modernization - or lets call it.. "Ageing". Elon Musk stated that the very reusability of Shuttle prevented from evolving at the same rate as other rockets. There were very few real design changes. And most of those were electronics and the material that the ET is made out of. At one point NASA even searched for components for reparation of it's aged computers that no one else made anymore, on eBay. Thus reusability, the holy grail of spaceflight for so long, ended up ironically stunting its own development.

Limited range - It was designed for LEO and possibly GEO, but not further. Even if you wanted, there is no reason to send a 50 tonne ship (that can carry 20 tonne cargo to LEO) any further.

Understated risk - At the beginning of the program it was assumed that the risk of failure is as low as 1 in 7000 flights, it was only realized, after the Challanger disaster that the risk was greatly understated and the risk assessment gave a possibility of failure of high up as 1 in 9 flights! (by looking into procedures of previous flights). Post Columbia safety improvements cut the failure chance to about 1 in 100, but it gave a huge cost and delays between launches, especially when you look that the Shuttle was predicted to have 50 flights per year (yet it averaged at 4).

Non-reusable ET - The external tank was not reusable and therefore had to be produced for each flight.

Please help me finish the list, or just discuss, or just blame me what I got wrong =)

Edited May 19, 2013 by nothke

[softweir]

I seem to remember that the orbiter had cooling system limitations that meant it had to spend at least 30% (approx) of each orbit in Earth's shadow. This imposed restrictions on the range of orbits available to the ISS.

My google-fu is failing me and I can't find any references to this. Maybe somebody else knows or can find it!

[Canopus]

It could also only stay in orbit for a short time due to the use of fuel cells.

[DerekL1963]

Flexibility - The shuttle was designed according to requirements set by the military. The crucial factor in the size and shape of the Shuttle Orbiter was the requirement that it be able to accommodate the largest planned commercial and military satellites, and have over 1,000 mile cross-range recovery range to meet the requirement for classified USAF missions for a once-around abort from a launch to a polar orbit.

This is an oft repeated myth that needs to die in fire.

The idea of the Shuttle started with something like a re-useable combination of Soyuz and Progress - a few passengers, a little cargo, shuttling back and forth between the ground and a space station. But over time, it gradually grew bigger and bigger - both because bigger is more flexible, and because bigger allowed independent flight. The increasing need for independent flight came from the realization the Saturn V wasn't coming back to launch a station in one go and the budget wouldn't stretch to cover a stream of lesser launches to provide for a smaller station. The same goes for cross range, it steadily increased over time because it increased safety. (Both by opening up additional abort options, and by providing more frequent and wider landing windows.)

The DoD requirements set the final performance requirements, but NASA's design wasn't so far off those... which is why it could be so quickly and easily adapted to DoD requirements.

For comparison, the heat shield of a typical pod manned rocket was composed of one single ablative heat shield

In the same way an aircraft's skin consists of one single layer of aluminum... I.E. it sounds simple when phrased that way, but in reality the heat shielding of the Apollo was a complex beast. It varied in thickness over the surface of the vehicle, it had to accommodate the hatch and the 'notches' for the windows, and the upper surface (I.E. that not protected by the SM) had additional insulation on top of it. As for much of the rest of the apparent complexity - it's a consequence of reuseability. (An ICBM requires much less maintenance than a jet fighter for the same reason.) You can streamline the required maintenance by using LRU's and such like, or learning more about what maintenance is and isn't required*, but you can't eliminate it altogether.

Lack of modernization

Was a consequence of budget - not only for modernization, but for a follow on vehicle. Apollo had the same problem when re-using the lunar CSM in the (very budget limited) Skylab program.

There were very few real design changes. And most of those were electronics and the material that the ET is made out of.

Aside from the big and very visible (and well covered in the media as a consequence) changes (like the engines, electronics, and ET), there were a whole raft load of detail changes over the years. One not well known is the steady changes to the TPS to reduce weight and maintenance.

At one point NASA even searched for components for reparation of it's aged computers that no one else made anymore, on eBay.

That was for ground support equipment, not flight equipment. NASA's infrastructure budget was never quite enough to cover it's needs.

* This could be seen in the Shuttle program where they went from removing and tearing down the SSME's every mission to borescoping them every other mission or so, removing them every third mission or so, and tearing them down evert fifth mission or so.

[Temstar]

Reusability

Reusability is in fact a con not a pro. It's much cheaper to build simple rockets and engines that is used once and thrown away than it is to engineer extremely complex engines and spacecrafts that has to survived to be used over and over again. By mass producing large number of simple rockets the cost per unit drastically goes down. The Soyuz/R-7 rocket family for example is quite inefficient and employs many archaic features such as using vernier engines for steering. But with so many launches per year over 60 years Russia can roll them off the production line like sausages if they want to and so on a $/kg to orbit basis they cost much less than extremely complex machines like the shuttle. Hence Soyuz is still going strong while the shuttle is already retired.

Flexibility

Flexibility is another con. Flexibility comes at a cost, and if you never use that 'flexibility' than it's worse than having no such capability. Case in point shuttle had the requirement to launch air force spy satellite into polar orbit from Vandenberg and then land back at the base after one orbit around. This required extreme amount of cross range capability which when implemented cost the shuttle in other areas. As it turned out the Air Force then changed their mind and end up using ELV for polar orbiting satellites so that cross range capability on the shuttle that costed a bomb to put on was literately not used even once.

[drakesdoom]

Reusability is in fact a con not a pro. It's much cheaper to build simple rockets and engines that is used once and thrown away than it is to engineer extremely complex engines and spacecrafts that has to survived to be used over and over again.

While this is true in the short term we cannot ever hope to maintain or expand a presence in space unless we can make reusable spacecraft. Ideally we separate atmospheric launch vehicles and everything else and build the everything else in space to stay there forever. There will be a transition period of awkwardness where we need to launch machinery into space from the surface, but once we get a space-born shipyard and asteroid mining setup there should be no reason to send industrial sized cargo up anymore.

[UmbralRaptor]

Cons:

* Recycleable, rather than actually reusable. (The SRBs and TPS in particular required serious refits after each launch)

* Low flight rate (135 launches over 30 years yields 4.5 per year. The goal was... substantially higher, and while this is better than the Soyuz capsule, it's rather worse than some expendable rockets.)

* Very expensive (related in part to the previous 2 issues)

edit: The large payload to ground capacity was never used. (Tying in with earlier posts on flexibility)

Edited May 18, 2013 by UmbralRaptor

[Fractal_UK]

Reusability is in fact a con not a pro. It's much cheaper to build simple rockets and engines that is used once and thrown away than it is to engineer extremely complex engines and spacecrafts that has to survived to be used over and over again. By mass producing large number of simple rockets the cost per unit drastically goes down. The Soyuz/R-7 rocket family for example is quite inefficient and employs many archaic features such as using vernier engines for steering. But with so many launches per year over 60 years Russia can roll them off the production line like sausages if they want to and so on a $/kg to orbit basis they cost much less than extremely complex machines like the shuttle. Hence Soyuz is still going strong while the shuttle is already retired.

Flexibility is another con. Flexibility comes at a cost, and if you never use that 'flexibility' than it's worse than having no such capability. Case in point shuttle had the requirement to launch air force spy satellite into polar orbit from Vandenberg and then land back at the base after one orbit around. This required extreme amount of cross range capability which when implemented cost the shuttle in other areas. As it turned out the Air Force then changed their mind and end up using ELV for polar orbiting satellites so that cross range capability on the shuttle that costed a bomb to put on was literately not used even once.

You very definitely need to prefix these statements with an "in certain cases" since as it is, what you have written could be interpretted as an argument that reusability and flexibility are undesirable in general, while the information you provide only supports that idea in certain circumstances.

The fact that we have no highly cost effective reuseable rocket at the moment is in no way an implication that we will never have one or that such a thing is impossible to produce. The SABRE engine is one example of a development that might fulful this criteria and offers the realistic possibility of a drastic reduction of launch costs for certain applications. Indeed most plausible engine designs that offer both reasonable TWR for launch applications and high specific impulse fall very definitely into the expensive to manufacture bracket, of course, if they can be used reliably for many launches, the cost of manufacture becomes increasingly negligable. On the other hand, conventional rockets are not going to budge too much from current Isp values and there is a limit to how much mass production can actually reduce costs.

The same is true with regard to flexibility and indeed this statement risks contradicting your first. If your rocket has little or no flexibility, how many times is it actually going to be used? If a conventional rocket cannot be used for a wide variety of mission profiles then the benefits you gain from having something that is mass-produceable are completely destroyed. There is no point mass producing components that will only be used once or twice. Rocket design is also expensive. You can spend years developing a wide variety of rockets for different mission profiles, testing them all, ensuring that your customer's cargo isn't going to explode on the way into orbit or you can design one or two highly capable rockets. You may not use the capability all of the time but providing you don't screw up and engineer something that's never or almost never going to be used, chances are you will see an efficiency gain from it.

[DerekL1963]

Reusability is in fact a con not a pro. It's much cheaper to build simple rockets and engines that is used once and thrown away than it is to engineer extremely complex engines and spacecrafts that has to survived to be used over and over again.

That's why we routinely throw away everything, rather than making it reusable... oh, wait. The proper response to high capital costs is to do what airliner and tanker owners do - use the heck out of them for as long as possible to amortize that cost over as many trips as possible. Not that high capital costs were the Shuttles problem, nor operating costs. It was the huge fixed costs that dominated the budget. But, just like airports and highways, the trick is to amortize that cost over as many flights as possible.

Cheap access to space, expendable or reusable is a bean counter problem as much as it is an engineering problem. One of the things that's held us back so long is the refusal of the engineer dominated space program to face that fact. The other thing is our collective unwillingness to actually spend the money to climb the ladder - it's going to take a lot of [design and operating] generations to work everything out.

By mass producing large number of simple rockets the cost per unit drastically goes down. The Soyuz/R-7 rocket family for example is quite inefficient and employs many archaic features such as using vernier engines for steering. But with so many launches per year over 60 years Russia can roll them off the production line like sausages if they want to and so on a $/kg to orbit basis they cost much less than extremely complex machines like the shuttle. Hence Soyuz is still going strong while the shuttle is already retired.

The number of launches over the last sixty years is no indication of the ability to mass produce them. (Or to put it another way - the Soyuz has flown fewer flights than the Shuttle... so the Shuttle should be amenable to mass production by the same logic.) Also, "mass production" isn't a magic wand that brings costs down - because what brings costs down isn't making a lot of things, it's reducing the costs per unit. There are many different ways to do that (some of which are collectively known as mass production), and it's not at all clear they're applicable to a complex vehicle like Soyuz. (And it's not at all clear what a Soyuz actually costs - we know what they charge per seat, but we don't know how they account for that charge or even if it bears any relationship to reality.)

Nor is there any evidence they can "roll them off the production line like sausages". (To be fair, there's no evidence the other way except the fact that they never have.)

Flexibility is another con. Flexibility comes at a cost, and if you never use that 'flexibility' than it's worse than having no such capability. Case in point shuttle had the requirement to launch air force spy satellite into polar orbit from Vandenberg and then land back at the base after one orbit around. This required extreme amount of cross range capability which when implemented cost the shuttle in other areas. As it turned out the Air Force then changed their mind and end up using ELV for polar orbiting satellites so that cross range capability on the shuttle that costed a bomb to put on was literately not used even once.

On the contrary - the shuttle's cross range capability was used on many flights, a straight in approach to the runway was actually a rare thing. As I said above, crossrange makes the vehicle safer by increasing abort opportunities and widening landing windows. As a result, NASA kept increasing the cross range as Shuttle development proceeded - even before the DoD briefly came onboard. I wish I could find the chart we used for input data, but we looked at this back around STS-100 and found that, on average, the Shuttle used about 40% of it's cross range capability - and the record was near 80%. (And that was a routine landing, not anything particularly urgent other than wanting to land it at the Cape to minimize turn around time..)

[DerekL1963]

The large payload to ground capacity was never used.

Other, of course, than all the times it was. (Like Spacehab, Spacelab, LDEF...)

[UmbralRaptor]

YThe same is true with regard to flexibility and indeed this statement risks contradicting your first. If your rocket has little or no flexibility, how many times is it actually going to be used?

IIRC, A single profile (launching into GTO) covers ~2/3 of all commercial flight. Though really, most expendable systems offer greater flexibility than the shuttle did. (eg: you liquid fueled upper stages are an option, launch directly into GTO or even Earth-escape, etc.)

Other, of course, than all the times it was. (Like Spacehab, Spacelab, LDEF...)

...Hubble.

[Fractal_UK]

IIRC, A single profile (launching into GTO) covers ~2/3 of all commercial flight. Though really, most expendable systems offer greater flexibility than the shuttle did. (eg: you liquid fueled upper stages are an option, launch directly into GTO or even Earth-escape, etc.)

That's fine but what kind of orbit you can deliver a payload to doesn't really describe the capabilities of your craft, an Antares rocket and a Saturn V can both deliver a payload to GTO but it doesn't really describe the respective capabilities of those craft. Payload to GTO is more meaningful.

Even then, the orbit your spacecraft can achieve goes a long way short of describing the capabilities that it might offer. Is it capable of orbital rendezvous which might make it useful for ISS resupply, for example? Is it human rated? Is it capable of re-entry? Can it return equipment to the ground? Or people?

[Fuzzy Dunlop]

The Shuttle's fundemental problem seems to be that it was, essentially, a space station which could land and be relaunched. While that is a brilliant technical achievement, it completely misses the point of a space station.

[Rune]

I'd say it was more a problem of using a Saturn-class launcher to put an EELV-class payload in orbit. And that completely defeated its intended purpose of cheap frequent launches. If the shuttle had become a 25mT reusable launcher with 25mT reusable crew and cargo options, it might have been what it was meant to be. Instead it was a partially refurbishable behemoth with an awful flight rate, payload, and crew safety. However brilliant the engineering behind the shuttle was, the design concept was flawed from the start.

Rune. Probably because a politician did it.

[BossSquirrelz]

I wish people would stop throwing their opinions around. Pros and cons are subjective.

But here's my opinion:

NASA's budget has been falling ever since Apollo. Don't you guys count that in? NASA didn't develop a new spacecraft, or improve significantly because they didn't have billions of $$$$$

America doesn't give two and a half s**ts about what happens in space. Anymore. We are like goldfish in a fish tank. Only instead of fish, there are corrupt, fundamentalist politicians.

[DerekL1963]

I wish people would stop throwing their opinions around. Pros and cons are subjective.

Certainly. But just because they're subjective, that doesn't mean they can't be based on facts. (But understanding of the facts is pretty thin on the ground.)

But here's my opinion: NASA's budget has been falling ever since Apollo. Don't you guys count that in?

I did - right in the first paragraph of my first reply. Much of the problems with the Shuttle stems from the decision (by Congress) to cancel the Saturn V (to limit the lunar program and save money*) and the space station, thus forcing the Shuttle to grow into a 'one-size-fits-all' mini space station. But that's not the end of the budgetary woes - because of the budget cap, NASA again and again was forced to chose what cost the least to develop rather than what cost the least to operate.

That NASA tried to leap from nothing to a fully operational Shuttle in a single generation didn't help any...

* Few people realize that the Apollo program was essentially cancelled in 1967 when Congress slashed the budget and capped hardware production. By the time of Apollo 11, the program was already running on fumes.