Hyperloop costs and challenges undermine pinned hopes | Hacker News Reader

Hyperloop costs and challenges undermine pinned hopes

bbc.com

60 pointsPoobuh83 comments

How's the Dubai to Abu Dhabi Hyperloop coming? That's the ideal situation. Flat route across desert. Government cooperation. Parties willing to pay for an expensive luxury system. If that doesn't work out, nothing will.

I agree. An evacuated tube build in a flat area with little development between the source and destination would really be a great proof of concept if they could get it built and running.

Temperature differential of day/night cycles there would be a bit of a nightmare though.

"But we are going to put it underground to deal with that!"

That's not the plan. Here's the promotional video for the Dubai project.[1]

[1] https://www.youtube.com/watch?v=3fu-6IDp3Fo

Well, I'm sure they might change plans. IIRC, the hyper loop videos in the beginning showed it to be all above ground.

Of course I have serious doubts that this would be a good means of transportation.

Real tunnels are much more complicated than pretty pictures of tunnels. Many tunnels have to hold up what's above them; they're really long arch bridges. Brunel's Thames tunnel looks like one.[1] Modern tunnels are usually round tubes, but structurally still arches.

Water is a big problem. On a good day, the New York City subway system pumps out 13 million gallons. There are bad days. Many cities are near sea or river level, so plan on a lot of pumping for urban tunnels. Water during construction has been the cause of many tunneling disasters. Some of Japan's longer tunnels hit underground rivers. Tunneling in soft wet soil is mostly about keeping water out.

Tunnels which carry large numbers of people need extensive rescue facilities. Usually, there are two tunnels, with cross-connections and emergency access points. Fire suppression and smoke control systems are needed. Eurotunnel has had five fires, three of them major, despite elaborate fire control systems and much advance preparation for emergencies.

The various Hyperloop documents gloss over all of these real-world issues.

[1] https://en.wikipedia.org/wiki/Thames_Tunnel

The biggest problem I have had isn't with the tech but the company.

There are many reasons Elon’s hyperloop will be in a tunnel, including safety, travel within the city center, and surprisingly much easier approvals. Utilities drill tunnels all the time, and so far for The Boring Company, getting approvals has been a matter of filing for a utility approval. There’s just so much less contention underground.

Yes cost is an challenge, but so much easier to focus on one problem rather than to be faced with many, like an above ground hyperloop would have.

“Without tunnels we’ll be in traffic hell forever” -Elon

Exactly. Building landtunnels uses much more material. This material is expensive; a 100KM+ metal twin tube including foundation is fundamentally more expensive then a railway.

Further, it seems to me the thermal expansion problem (to me the single biggest technical problem) is not an issue either.

No wonder Elon Musk started The Boring Company. Once costs go significantly down to bore tunnels, most land on earth can be used to build a highspeed network on.

> There’s just so much less contention underground.

Except for all the utility conduits (only half of which are actually marked), gas pipes, steam pipes, water pipes, sewage pipes, forgotten tunnels (remember the Chicago flood?), subway tunnels, basements, foundations, and pilings. Sure, you could avoid a lot of that by going very deep, but you still have problems with ventilation and access shafts to deep tunnels.

> “Without tunnels we’ll be in traffic hell forever” -Elon

The way to fix traffic hell is to accept that, if you live in a large city, space is at a premium, and your single-occupancy vehicle takes up 20 times the space as standing on a train.

People obviously prefer SOVs to standing on trains. Why not give people what they want if we can develop technology to do it?

The entire conversation surrounding city planning, whether the specific topic is transportation, zoning, or utilities, has take on a weird ascetic moralizing tone, where the subtext is that the modes of existence people choose for themselves is somehow "bad", that we need to accept limits, and that the accommodations and compromises consequently made are good for us.

I don't share that mindset. To borrow an old phrase, better living through technology really is possible.

Where good public transport systems exist, people use the hell out of them. It is so much nicer to travel by train from point-to-point and not have to worry about parking.

What if parking is done by the car via self driving technology? Would you still prefer the train?

In rush hour in plenty of cities, trains will almost always be quicker, because they can move more people per hour than the road network can. (If you look at some metro systems, you can get up to 60,000 passengers per hour on a single track; for a highway, 2,000 cars per lane per hour is considered good: if you're all in SOVs, you're talking 30 lanes in each direction to be comparable with a single train line.)

What if the 'train' picks you up at the door and gets to your destination exactly when expected? would you still prefer to park autonomously blocks away?

At least where I live, parking is sparse and therefore expensive. That's a good thing, because parking space is a waste of land that could be put to a more productive use.

So yes I still prefer the train. Buses are cool too!

I prefer solitude. Give me a compartment on that train and I'll reconsider my preference for automobile transportation.

put on some headphones and a blindfold then, cars really don't work in high-density environments. if you want to be alone, try living somewhere with less people where cars actually make sense.

Well, a city is about living together in a community. "Bad" lifestyle choices certainly exist, such as littering - it's something that can be helpful to the individual but have a negative impact on the community. Planning a city is as much about shaping cultural values as it is about adapting to existing ones.

> Planning a city is as much about shaping cultural values as it is about adapting to existing ones.

No it isn't, and city planners employed by the people shouldn't try to use their influence to shape culture.

There is no realistic apportioning of resources, or set of plausible new technologies, that will give safe and effective cars the >7 billion people on this planet in the next century. At least not without drastic downsides. Cars work great for the wealthy, and are tolerated by the less-wealthy for lack of better options, but are not in any circumstances an egalitarian option. (And, btw, that's likely why you feel like the arguments against cars have an "ascetic moralizing tone"- use of single-person cars is a taken-for-granted habit into which is built deep inequality.)

I don't think you're right about car technology scaling, but let's suppose you're right. Why it would follow from cars being infeasible for seven billion people that the citizens of a rich first-world country should have their car use discouraged?

Radical egalitarianism is an intensely political position, and city planners shouldn't be pushing it on the population at large by artificially discouraging certain modes of transportation. They should focus on the preferences of the populations they actually serve, not on some Utopian ideal.

I personally have no problems with people taking their SOVs into the cities, but you do have to acknowledge that taking up excessive space obligates you to have to pay extra costs for taking up that space. In practice, many people who want to commute via SOV often take great umbrage anytime a city takes steps to make them pay the true cost of that decision.

There's nothing particularly "radical" about it; if you weight everyone's utility equally when you design, e.g. a city, then the result is one that is definitionally egalitarian. And that is precisely what it means for governments to, as you say, "focus on the preferences of the populations they actually serve." Egalitarianism is only "radical" in the sense that, in most real contexts, some the desires of some subset of people are considered more important- for instance, the people who can pay more, or are a more effective voter bloc.

Not that I think the above is the decisive factor against wider adoption of public transportation. The bigger issue is an activation energy problem- public transit is only bearable and useful when a substantial portion of public funds are allocated to them, but the bulk of the public isn't currently inclined to have that happen since they don't presently use the (unbearable and useless) public transit option. Cars are a fairly vicious local maxima.

> There is no realistic apportioning of resources, or set of plausible new technologies, that will give safe and effective cars the >7 billion people on this planet in the next century. At least not without drastic downsides.

What are you basing this on, and why would you say it as if it is obvious? This is exactly the sort of weird fatalism the GP is talking about.

There are already a billion cars on the planet, which is enough for 7 billion people with autonomous self-driving. 90% of US households own a car and the rest of the world will catch up with US wealth within a century. Even current-day-tech electric cars and nuclear power could be used to power them without carbon (at large expense but one that could be afforded after a century of economic growth).

People don't know what they want, that is why advertising is consumer farming.

> People obviously prefer SOVs to standing on trains. Why not give people what they want if we can develop technology to do it?

SOVs just take up so much space, about 20× that of normal standing density. In cities were even the mass transit is oversubscribed, there is no possible way to fit in SOV. Even burrowing in tunnels doesn't work--you'd still have to somehow squeeze the access ramps in. Cities that try to be SOV-centric instead of transit-centric end up having really large sprawl and lose the benefits of being close to anything.

To give you a sharp visual statement of the difference in space. When trains disgorged their passengers at Union Station in Toronto, people would often stream across the surrounding streets without particularly regard for crosswalks. When traffic was waiting at the light on Simcoe St, there was more people walking between the cars than there was sitting in the cars waiting for the light to change.

In places like London, a major problem with the underground system is that even space to surface vents for aircon etc. is restricted enough for it to be a problem. Fitting access ramps would be a massive problem.

The lack of space for aircon vents has gotten to the point where the clay surrounding the underground tunnels are permanently 5+ degrees warmer than it was when the system was built, because not enough heat is pumped back out, and it soaks up less and less of the waste heat. Most of central London is basically sitting on top of a giant clay heatsink.

Building down also doesn't add all that much when trains run frequently enough that junctions/intersections are not viable without delays or sacrificing speed. As well as when depth changes need to be very gradual...

The irony is that building more tunnels is by far easiest in cities too small to need the capacity - if you have long intervals between trains, for example, junctions become a far more manageable problem, leaving far more trajectories free to use.

A lot of human activity has positive and negative consequences that have to be balanced. We have a democratic system that gives some of that responsibility to government.

With this kind of debate you will hear a stream of fact based arguments about the correct thing to do. Arguments that are put in opposition to one another. Is it more important to reduce journey times or save a particular habitat? Is the view out of your window more important than reducing emissions by building wind farms. These kind of questions are impossible to answer in an objective way because it is so ridiculous to create such dichotomies.

The facts do matter, but in the end it always comes down to a moral judgement. That is how humans balance complex conflicting issues. And this applies just as much to technology as anything else. By believing in the promise of technology you are making a moral judgement. But it is far too complex to really understand fully. That is why you make a gut decision.

> Except for all the utility conduits (only half of which are actually marked)

You dont have these in Arabia. All you have at 30m depth is desert.

subway tunnels

Well, that's the point, isn't it? These are like subway tunnels, which are a solved problem and implemented for decades - including multi-level tunnels. Musk just seems to want to make them cheaper and put some space-age motorail inside it.

This doesn't achieve the density of subways, but it isn't meant to. In the cases of low-density areas, it solves the last-mile problem by bringing the vehicle with the person on the sled. In the cases where it's high-density to high-density, the sleds can function as subway cars.

The question in my mind is more of whether the perceived benefits from the improvements that Elon is planning in the process can be realized, and if they'll be enough.

The equation changes a bit when you take into account that you can only have one train per X miles of track, while you can have many cars per mile.

It's important to note that evacuated tubes aren't a new idea. The Hyperloop is a re-branding of Evacuated Tube Transport (ETT) that came about in the .. 80s? I think? I wonder if Musk is licensing the patents from that company or if the Hyperloop division has patented their own tech as being entirely different somehow.

Edit: Looks like Musk did negotiate with the ET3 people back in 2013

https://en.wikipedia.org/wiki/ET3_Global_Alliance#2008–prese...

Strange that the hyperloop idea is to have tubes above ground, whereas the Boring company is all about tunnelling under cities.

Not that strange. Can't never get enough tubes. Underground, above ground, on mars. Before long we'll be talking about tubes in space; what's hype-er than hyper?

I mean, evacuated tubes are ultimately a development of atmospheric railways, which date back to the 18th century.

There were definitely lots of proposals to put maglevs in evacuated tubes from not long after large amounts of funding started to pour into maglev design.

Hyperloop is a variation on the Vactrain [1], a concept that goes back a long time. The thing that made Hyperloop different was the notion that you didn't need a hard vacuum in the tube -- the air in a Hyperloop tube will be really thin, but it will be there. The other thing notable about Musk's original proposal was that the cars would float on air bearings rather than a maglev track. I think most if not all of the current experimental Hyperloop systems are using maglev technology because it works better than the air bearings would. If Hyperloop systems adopt hard vacuum in the tubes, then I think you can make the case that Hyperloop is not new, it's just re-branded Vactrain. In reality, I don't know that this distinction will make any big difference.

I personally think we will ultimately see some functional passenger-carrying Hyperloop systems, but I don't know that they will be any more successful than maglev has been to date. I don't think we will see any systems using a hard vacuum (the classic Vactrain), because hard vacuums are, well, really hard to do.

[1] https://en.wikipedia.org/wiki/Vactrain

I'm curious, what's the definition of hard vacuum?

Obviously there will be a few nitrogen molecules bouncing around in there so it seems there must be some arbitrary cutoff.

I worked with electron microscopes back in the day, for that I would say 10^-6 atmospheres. But it probably depends somewhat on your application, I don't think there is a solid definition.

There’s pretty much no such thing as a ‘hard vacuum’ in the real world. Even ultra-high vacuum systems always have particles still floating around (even space has a few hydrogen atoms per cubic metre, while the best vacuums we can produce on earth have a few orders of magnitude more particles).

At 100Pa (according to Wikipedia), the hyperloop is still a bit less than 0.1% of the pressure of the atmosphere, which is still a significant challenge given its size, expansion problems, and the risk of a small impact destroying the whole thing. The amount of energy required to pump down a hyperloop of any reasonable length (which could take days) would be immense.

I think with the expansion and impact risk, the only safe place for a hyperloop would have to be underground. Even then, I think it’s unlikely to ever really be practical.

British scientist Philip E. Mason (thunderfoot) made several skeptical videos on Hyperloop. https://www.youtube.com/watch?v=RNFesa01llk I highly recommend his whole Busted! series.

thunderfoot's analysis forgets, among other things, that the force of pressure scales with r^2 and mass with r^3

Building a mechanical system isn't the same thing as writing an algorithm. You can determine ahead of time r and p for all uses, and m has a pretty manageable upper bound too. Asymptotic considerations are not always relevant. They do have real utility in academic work, where a designer of a proof of concept ought to show that the applicable cases of his idea are workable. But that's not really the situation here.

More importantly, maximum velocity is not likely to be a strong function of mass, which is relevant only to acceleration and friction. Acceleration/deceleration can be assisted by other means, and bearing designers can adopt very effective solutions from other means of transit.

I'm specifically referring to why the demonstration of a ball bearing puncturing a glass tube does not scale catastrophically at larger sizes.

I have a lot of trouble trusting Thunderfoot's judgement after he waded into the Sarkeesian thing.

What will the actual humans-per-hour throughput of a practical hyperloop implementation? My understanding is that although the technology may work, it's unlikely to be economically feasible given the actual realistic capacity. How is this superior to high-speed trains?

Seems to be the big problem with this. I can't see it being anywhere near a high speed train. The distances between pods would have to be (very) significant and the pods themselves will be pretty small I bet as it's in a pressurised tunnel.

Other problems include branches and junctions which normal railways can do with no problem.

The throughput of hyperloop is equal to ~3-4 trains per hour of capacity, assuming you completely maximize utilization of the infrastructure (that's ~30s dispatch intervals). This is assuming you could safely dispatch pods at 30s intervals, which is a contentious issue to say the least. For HSR, you'd generally see 6-12 tph for intercity trains, although the tracks themselves can generally handle 24-30tph (depending on how good your signalling is).

Effectively, hyperloop is the capacity of a point-to-point transit system (e.g., airline) with the infrastructure requirements of a train network, or combining the worst features of both systems. You'd have to bore city-center tunnels for every city you'd want to connect to, as opposed to being able to share those tracks (the fact that you can't share intercity rail with commuter rail is a big reason why maglev generally doesn't make economic sense).

It's completely dependent on how long the cars are. You could do hyperloop* with 1,000 passenger trains just fine. The real limit is utilization, and smaller cars are easier to design.

The real benifit is simply speed, it's faster than aircraft and no need for extensive security as you can't hijack trains.

* Not that the term seems to refer to any specific design at this point.

I was going by the original Hyperloop concept as opposed to vactrains. It can sometimes be hard to tell if someone means "Elon Musk's 21st century idea" or "yet another retread of a 19th century proposal" when Hyperloop is mentioned, but I tend to err on the side of the former, since most of those adherents are extremely adamant that the differences means it's better.

You're right in that vactrains are essentially no different than HSR, and that's basically because vactrains are just ultra-HSR. There is some issue about using maglev propulsion versus regular electric locomotive, but that's mostly about being able to share infrastructure in inner-city portions.

A bomb is going to be just as horrifying on a Hyperloop as it would be on a plane - killing hundreds and destroying even more infrastructure then a bomb on a plane would. Security is still a need.

It does not need to be impervious, just no better a target than a sidewalk.

I don't think a reasonable sized bomb is going to do much more than kill people in a single car. So, sure some people would die and it would cause some damage, but not even close to hundreds. Basically, the same risks as someone blowing up a buss, which has no real safty checks.

If a bomb causes a breach of the loop, what happens to the other cars? If cars are going rapidfire, what happens to the next car if one car stops faster than braking speed, or causes a deformity of the tunnel that obstructs it?

You certainly can hijack trains [1]. If you plan on trying to get away, it's obviously not a very smart choice.

[1] https://en.wikipedia.org/wiki/1977_Dutch_train_hijacking

Even if it’s respectably higher than anything existing, it can only source passengers from hub origins and drop them at hub destinations, leaving last-mile problems yet to be solved. Spokes will be needed at destination hubs, that allow for individualized selective destinations for the last mile.

Meanwhile if a city needs 3/4 of it’s workforce through the door by 9AM, congestion at each hub still occurs, and rush hours aren’t actually alleviated, only spread around. Hub stations will be the same horrible kind of place train stations, bus stations and airports are, unless a new conceptualization of origin sourcing and destination arrival is brought to the table.

The only new aspect that a hyperloop adds to the equation, to change the game is that, theoretiacally, longer distances can be negotiated, permitting any given urban center to import from greater distances than before, and urban residents can interchange between hubs without 500MPH aviation.

Of course, for those promises to hold true, tubes must be cut, and track laid within them. Every mile of interconnection adds compelling power to the principle of the concept. Individual point-to-point connections are meaningless, if there are not many points connected, and the distances aren’t longer than was previously realistic.

Hyperloop ultimately is about a way to bring the advantages of high altitude jet travel to much shorter regional hops, essentially bringing the stratosphere down to sea level. It's not really about mass transit within a city. Don't need 500mph for that.

I first read about Vactrains in the Usborne Book of the Future of the seventies, which credited Dr Robert Salter, although the idea goes back a fair bit further.

https://en.m.wikipedia.org/wiki/Vactrain

Elon's idea seems to be to make it quite a bit smaller, and to have the tubes above ground, which does make it more feasible.

Is his association more to do with branding?

Would it be economically feasible to use the hyperloop to transport just goods and not people, similar to a pipeline vs train transport? I always see it referenced in connection to human travel, but it seems like safety is one of the biggest issues.

What’s unsafe about it? Seems a lot safer than a plane.

seems like a lot of people think compromising the vacuum could lead to train derailment and very likely the death of everyone on board (they're traveling at 700 mph after all)

It's far less at-risk if it's in a tunnel under the ground.

Actually, tunnels are riskier than at-grade. A fire in a tunnel is basically a deathtrap, and can even seriously impact people not inside the tunnel (see https://en.wikipedia.org/wiki/Howard_Street_Tunnel_fire for an example of the latter).

How would a fire get going in a near-vacuum?

There's potential issues with earthquakes and terrorism, both of which are less likely to affect planes, since they travel 10k from the ground. (Terrorism still obviously affects airplanes, but with Hyperloop you not only have to secure the stations, but the whole track too).

Hyperloop wouldn't work as an effective weapon like an airliner can, though.

Airliner-as-weapon was the "innovation" of the 9/11 attacks. That's one of the big reasons why the attacks worked, and why they'd never ever work again: previous hijackings were about using the plane + passengers as hostages, in order to achieve some other goal. So airline staff had a "script" of not risking lives to resist a hijack attempt, because hijackers typically didn't want to kill anyone, since killing people on the plane would reduce the number of hostages available to use later as leverage. The plan was to let them get the plane on the ground, then wait for the Delta Force guys to show up and deal with the situation. When that changed in the 9/11 attacks, you suddenly saw people on planes actively resist hijacking even when it risked lives (in fact, that started happening on 9/11, with the United 93 passengers once they found out what had been done with the other hijacked planes).

Meanwhile, people remain afraid of mass-casualty incidents, and so they've remained an effective tactic (there have been several subway/train attacks in the post-9/11 era) even if nobody managed to crash the trains into buildings.

There has been plenty of hijackings since 9/11, and only in a few instances have passengers tried to resist. Maybe things would be different in the US, but I suspect even that would be highly contingent on how the hijackers would act (e.g. whether they'd have the presence of mind to make a big point of stating an intent that'd involve keeping everyone alive).

EDIT: It's worth noting that hijackers have overall been far more likely to get killed during a hijacking than passenger - as a passenger, unless you have very specific reasons to believe they're intending to crash the plane, it seems like a much safer bet to keep your head down and wait.

If those are the biggest risks then sign me up. I take bigger risks crossing the road every day.

> with Hyperloop you not only have to secure the stations, but the whole track too

I hear this argument a lot, and it personally doesn’t make sense to me. Trains tracks in the middle of nowhere have functioned for two centuries.

And trains have been attacked for at least 150 years or so (westerns, Lawrence of Arabia, https://en.wikipedia.org/wiki/Cairo–Haifa_train_bombings_194...), just like stage coaches got attacked for centuries before.

On the other hand, technological progress means planes nowadays also can be attacked while in air, using rockets (https://en.wikipedia.org/wiki/Malaysia_Airlines_Flight_17) or (not quite yet, I think), drones.

Planes have an advantage for the defense that it is easier for them to randomly change routes.

Air travel is extremely safe, though. In fact, we went all 2017 without any fatalities from air travel.

That’s partially not true. 2017 was the safest year, yes, but there were still fatalities. The first page[0] on Wikipedia’s category for 2017 aviation accidents[1] was a flight where all 8 people died.

[0]: https://en.wikipedia.org/wiki/2017_Aerogaviota_Antonov_An-26...

[1]: https://en.wikipedia.org/wiki/Category:Aviation_accidents_an...

Does anyone have any idea what happens if a capsule breaks down in a tube. Crawling along an evacuated vacuum tube for scores of km doesn't sound enticing. After all lifts breakdown now, and they don't go that far.

Kind of like asking what happens if an aircraft breaks down during flight. It's very bad, so lots of engineering has to go into making sure it never happens.

Also, I agree with the article:

> launch a commercial operation by 2021

is an insanely optimistic goal.

The article didn't seem to really discuss any limitation other than the general and unspecific incredulity of the authors.

right?!

Exactly. He quotes engineer after engineer saying "we've already proven it's feasible" (and not a single one opposing!), and follows each quote with something like "bringing him back to reality, I asked him...".

As a Brit that moved to SF, this kind of thinking is precisely the reason why.

Reply on news.ycombinator.com