Hyperloop. You saw the proposal. I saw it. Your grandmother saw it. It sounds great. Cheap, fast and from a guy with a lot of charisma. What.s not to like? In fact, it does sound like for someone who can develop a working spaceship, making a ground transport system would be a trivial task, and that might lull you into a sense of safety. Don.t let that happen. Spaceships fly once every few months, trains go daily for years, so if a fixed chance p of failure exists, the expected failures per year is highers of something used more often.
I think we can all agree that a fail-proof mechanical device cannot be created - you can thank thermodynamics for that. All things that move will eventually stop moving. Ignoring even that, all things made by humans break. All steel beams have weak points; all control software has bugs; all contractors building things sometimes do not meet the final specification; hell, even all clever designers proposing fast transport systems make mistakes. The point here is simple: no claim for a system that never fails is possible in our world. Why is this relevant?
The proposal calls for an 800pmh train riding through a tube positioned above ground over existing roads and highways. For comparison, your garden variety 737 flies at about 650 mph. At 800 mph, every kilogram of payload has just under 64 kilojoules of energy. This number is hard to put into context for many, so let.s do it a bit differently. A normal passenger train car weighs 20 tons empty. Let.s give Musk a great advantage - we.ll pretend he found a way to halve the weight somehow! Adding in 60 people at an average of 77 kilograms each, we get 14.62 tons.This train car, traveling at 800 mph thus has just under a billion joules of kinetic energy (935 million joules to be exact). Now, the context: this is equivalent to the energy of exploding 223 kilograms of TNT.
Back to our flawed world and imperfect engineers and contractors and steel beams. Eventually something will go wrong. There exists some chance that when it does, the train will be contained in the pipe, and will slowly decelerate. But, as we had mentioned before, steel beams and contractors make mistakes too. Wear and tear does not help. Eventually a train will leave the tube. And now we have a train, with the kinetic energy of 223 kilos of TNT over an existing road, or over a city center (where the hyperloop stop will be). Now, the thing you must understand about amounts of TNT that are measured in kilograms, is that such things never end well for anyone in the vicinity. Not only would everyone in the train be instantly killed, it is likely that a few buildings and countless people on the ground would not fare very well either. Even though the train will slow down before a city it stops in, it will pass over many others along the way at full speed, since it was supposed to be routed over existing highways. Highways, which, believe it or not, pass though small and medium sized towns. Even ignoring that, just crashing onto a highway will cause plenty of casualties on it.
One can attempt to argue that this can be solved with clever crumple zones, or bumpers. In reality this is not an option. As per the first law of thermodynamics, the energy cannot simply disappear, it must be transferred to another location or another form. Bumpers, for example, convert kinetic energy to thermal energy. Let.s try that. Let.s say the front of a train is a one ton steel bumper. Let.s even go further. We.ll say that it will absorb all the energy successfully. (None of this overly-optimistic scenario is possible, of course). What would that do? Well, it would rase said bumper.s temperature by about 2000 degrees celsius (2040, to be exact). Of course steel would not handle that and evaporate. This is all inconsequential, because such a bumper is not possible, and additionally this is one extra ton to add to the train weight, bringing even higher in the amount of energy we must dissipate.
You may be curious why I brought up a jet as a comparison in the beginning. As you may see, jets have similar problems (lots and lots of energy). How do jets solve them? Well, not always well (see Air France 447). But jets have one advantage over hyperloop - they are ten kilometers over the nearest hard surface. This gives them three things: 1. time to decide on a course of action and act, 2. altitude to glide to somewhere safe, 3. ability to dissipate the energy safely and slowly (gliding). These methods cannot be safely applied to hyperloop: (1) it is not far enough removed from the nearest hard surface, (2) it cannot glide, and (3) given the inertia involved, you would not be able to change its course fast enough due to (1).
Simply put, putting anything weighting in the tens of tons range and moving at speeds in the hundreds of kilometers anywhere near residential areas is a bad idea due to the reality-guaranteed non-zero failure rate, and the complete lack of a safe way to dissipate that much energy in the little time available. Thus the failure modes of such a transport are not tolerable anywhere in the vicinity of residential areas.