About the SA Blog Network



Opinion, arguments & analyses from the editors of Scientific American
Observations HomeAboutContact

We Could Make Elon Musk’s Hyperloop Real

The views expressed are those of the author and are not necessarily those of Scientific American.

Email   PrintPrint


Courtesy of Elon Musk

At some point cloaked in the (recent) mists of time, Elon Musk took over the mantle of leading U.S. visionary, as is the prerogative apparently of our technology billionaires. That’s in no small part because his current crop of companies—SolarCity, SpaceX and Tesla—all began to thrive. But it’s also because the South African native turned American entrepreneur has a notably offhand way of imagining a more plentiful—and beautiful—future. The man’s musings about something he called the Hyperloop last summer set off a frenzy of tech blogging and tweeting. Would Musk finally solve the problem of wildly unglamorous yet vital public transportation, as he seems to have solved the conundrums of the electric car, private spaceflight and cheap solar power? (Well, not entirely, but you know what I mean.)

On August 12, after a self-professed all-nighter, Musk unveiled just what in the heck he had been talking about as the Hyperloop—with a little part-time help from engineers at Tesla and SpaceX. It turns out to be a clever kludge of ideas that have been floating around since the 19th century—all designed to take people from Los Angeles to San Francisco in 35 minutes. Take a pod that would carry passengers all in a row and put it in a tube where it can be accelerated to high speeds like a bullet in a gun. Don’t make the tube a vacuum, however, since it’s all too easy to mess up a vacuum-based system given nature’s general abhorrence and whatnot. Nor should the tube be normal pressure either, like the pneumatic mail tubes of yesteryear, given that pushing all that air around would entail too much energy when expanded over hundreds of kilometers, not to mention an absurd amount of friction on the tube walls.

Musk’s solution was to cheat: make the pair of steel tubes low pressure, roughly 100 Pascals or “1/6 the pressure of the atmosphere of Mars,” he wrote in the proposal outlining the Hyperloop concept posted online. And, to get around that little problem of airflow, have the pods come equipped with a hole and a battery-powered fan and compressor so that the air in front is sucked in and then used as the traveling cushion—think air hockey—underneath the pod’s 28 skis. Oh, did I not mention the Inconel skis? These struts, made from a superalloy more commonly found on spaceships and nuclear power plant steam generators—which should give you some idea of the kinds of temperatures and pressures the supermetal can withstand—help the pod glide along the tube at 1,220 kilometers-per-hour, with a boost every 110 kilometers or so from the kind of linear induction motors Tesla uses to make its amazing cars go, flattened along the walls of the tube. Those motors, and the batteries aboard the pods, would be charged up by all the efficient photovoltaics laid out atop the tube’s roof, producing an estimated 57 megawatts-worth of power, or nearly three times as much as Musk calculates the Hyperloop would need to operate. “You’d have to dump the power somewhere,” Musk noted in a 30-minute conference call with reporters to discuss the big idea.

Top speed could be reached in as little as 35 seconds with no life-threatening gravitational pressures on the 28 or so passengers (possibly including three “full-sized” vehicles if we want to pay for the deluxe model.) The whole multibillion dollar tube would rest on at least 25,000 six-meter-tall pylons (with dampers inside to protect against earthquakes) placed every 30 meters or so that would follow the same route as I-5, keeping land and permit-buying costs to an estimated $1 billion. Total travel time would be 35 minutes between downtown L.A. and S.F. Elon calculates your ticket on the Hyperloop would set you back about $20, though each pod would cost around $1.35 million.

So far, so good, and no violations of the laws of physics, if possibly of the laws of economics. But there are some readily apparent flaws. Even at low pressure, there’s an awful lot of air resistance, a.k.a. friction, heating the air inside the tube, which is why each pod would need to carry some 800 kilograms of water on board strictly for cooling purposes (for those non-metric types that’s almost a ton)—adding weight and perhaps instability, with all that water sloshing around under fast acceleration. By the end of the trip, that water would be steam and, somehow, much like Better Place’s failed battery swap system, those ton-heavy cartridges of cooling water would be swapped out. That’s some feat, and we might want to consider other cooling fluids.

The air skis would also start to lose their luster at lower speeds, so the pods might need wheels for takeoff and landing. The linear accelerators lining the tube would need their own battery packs, in case the electricity got cut off, so that each pod already in the system could complete its journey—and it’s a little unclear (to put it kindly) what would happen if one of the pods suddenly depressurized inside this partial vacuum tube. The 1 g passenger experience also might not be for everyone (consider that roller coasters subject thrill-seekers to roughly 0.5 g), though most of that would be pressing down on the passengers as Musk noted in a conference call with reporters.

This hyper loopy concept springs from Musk’s disappointment in California’s high speed rail plans, which are both slower than other trains out there at an average speed of 264 kph and more expensive per mile at a total estimated bill of nearly $70 billion. Of course, that has something to do with California land prices and U.S. rules and regulations. The Hyperloop, such as it is, offers no solution to political gridlock, NIMBY-ism or high land prices. In fact, it is almost absurd to suggest that all the land that would be needed for the Hyperloop to connect L.A. and S.F. would cost just $1 billion. If the physical science of the Hyperloop is not outrageously loopy, there is a near complete vacuum on the social science side.

Ultimately, the Hyperloop could cost less than the very real world high speed rail proposal because it exists in a vacuum, a protective air cushion created by any engineering proposal that only exists on paper. I salute the innovative thinking but there’s also a lack of foresight, and let’s remember this is not a new idea. It is an old idea that, as even Musk admits, has never found a way to get built.

As it stands, no one actually plans to build the Hyperloop, not even Elon, though he might go so far as to fund a prototype. “I think I would have to punt it for a little bit of time,” Musk mused on the conference call. “It wouldn’t be immediate.” Maybe one of Elon’s superrich friends—Google’s Sergey Brin or his PayPal co-founder Peter Thiel—could divert some of their riches from other vanity projects like Brin’s lab meat or Thiel’s seasteading to fund development of the Musk-mobile. “If somebody else does the demo, that’ll be really awesome,” Musk added, suggesting if he’s the only one doing the Hyperloop, it would be pretty far down his list of tasks and might take at least a whole decade to accomplish.

That’s a bit sad given that one of Scientific American‘s original editors, Alfred Ely Beach, built a pneumatic tube train—in 1870, under New York City. No small feat. Surely we citizens of the 21st century could build a pneumatic tube train over the I-5 highway if we really wanted to, no?

Or we could just bring back the Concorde, or some other supersonic airplane and forget about the tubes. As Musk notes “with a high enough altitude and the right geometry, the sonic boom noise on the ground would be no louder than current airliners.” Maybe he’ll build that instead. Or, as he wrote of figuring out real teleportation: “someone please do this.”

David Biello About the Author: David Biello is the associate editor for environment and energy at Scientific American. Follow on Twitter @dbiello.

The views expressed are those of the author and are not necessarily those of Scientific American.

Rights & Permissions

Comments 25 Comments

Add Comment
  1. 1. Doc Who 10:39 am 08/14/2013

    A more practical idea would be to convert I-5 into an autobahn. California is replete with automobiles with top speeds in excess of 120 mph, whereas the speed limit on I-5 is a mere 70 mph.

    Obviously, many motorists cruise at 95 mph, resulting in a dangerous speed disparity.

    I-5 is virtually straight for hundreds of miles, so the increased speed limit of an autobahn makes perfect sense. Increasing the maximum speed limit to 155 mph would alleviate the excruciating boredom of a motoring along I-5.

    Naturally, the speed limit would be lowered under tule fog conditions, as it is now.

    Link to this
  2. 2. sault 11:29 am 08/14/2013


    Going at 120 mph, most vehicles are probably getting 10 miles a gallon or so. Completely wasteful. And when people DO get into accidents, the results will be horrific.

    Link to this
  3. 3. Scienceisnotagenda 1:21 pm 08/14/2013

    Unfortunately faster means of transport means more use of resources. Way more travel. First comment yesterday in the news….now you might live in city a and work in city b. Perhaps go shopping for the day, etc.

    Faster transport doesn’t just replace slower means but expands the usage pie. This is why high speed rail up is not so great for the environment. Patterns of social movement increase exponentially. A university course or sports event or shopping, etc. becomes accessible ‘elsewhere’.

    Link to this
  4. 4. plswinford 3:11 pm 08/14/2013

    People keep saying “nature abhors a vacuum”. The universe is almost entirely vacuum.

    Link to this
  5. 5. sethdiyal 5:25 pm 08/14/2013

    “. In fact, it is almost absurd to suggest that all the land that would be needed for the Hyperloop to connect L.A. and S.F. would cost just $1 billion. ”

    Biello has taken leave of his senses again. The project will be built on pylons in the center of I5 – public land – not so absurd.

    Link to this
  6. 6. And Then What? 7:25 pm 08/14/2013

    Nature will always demand a “balancing of her books”. You can have whatever she can supply so long as you are willing to pay her price. If you want to move mass numbers of people a high speed she will require that the total amount of energy be balanced. There is no free lunch in her Café.

    Link to this
  7. 7. Steve926 12:43 am 08/15/2013

    This seems like something Disney would be all over. The proof of concept could be their new ride for Tomorrow Land, then over charge everyone for the full scale trip.

    Also, instead of low pressure air tubes, can they use a lower density gas, like helium.

    Link to this
  8. 8. rkipling 3:43 am 08/15/2013


    Perfect for Disney! If you let some helium leak in, then everyone could talk like Mickey.

    Link to this
  9. 9. Satya Narayan Tiwary 9:07 am 08/15/2013

    Hyperloop concept of transportation is beautiful and wonderful idea. This idea must be converted into reality. This will add to the world.
    S. N. Tiwary

    Link to this
  10. 10. ScienceIsWinning 2:49 pm 08/15/2013

    First off, you really really need to distinguish what “readily apparent flaws” are from your analysis and what you’re basically reciting from the paper.

    “The air skis would also start to lose their luster at lower speeds, so the pods might need wheels for takeoff and landing.” – These wheels are not only part of the proposed design (pg. 20), but are later mentioned in sec 4.5.3 to function as an emergency safety feature to be able to propel the capsule under its own on-board power.

    “The linear accelerators lining the tube would need their own battery packs, in case the electricity got cut off, so that each pod already in the system could complete its journey” – Again, this is part of the proposed design (pg. 33)

    “and it’s a little unclear (to put it kindly) what would happen if one of the pods suddenly depressurized inside this partial vacuum tube.” – It’s a little unclear (to put it kindly) if you know how to read a table of contents and see that there is an entire section on pod depressurization. But I guess I’ll have to post it in detail …


    Hyperloop capsules will be designed to the highest safety standards and manufactured with extensive quality checks to ensure their integrity. In the event of a minor leak, the onboard environmental control system would maintain capsule pressure using the reserve air carried onboard for the short period of time it will take to reach the destination. In the case of a more significant depressurization, oxygen masks would be deployed as in airplanes. Once the capsule reached the destination safely it would be removed from service. Safety of the onboard air supply in Hyperloop would be very similar to
    aircraft, and can take advantage of decades of development in similar systems.

    In the unlikely event of a large scale capsule depressurization, other capsules in the tube would automatically begin emergency braking whilst the Hyperloop
    tube would undergo rapid re-pressurization along its entire length.”

    Basically all you offered was: maybe we should look at non-water coolants and 1g of downward (not lateral like the roller coasters) acceleration might be uncomfortable for some. Hardly real flaws for an alpha design.

    And finally: why in the world do you compare this to a pneumatic tube at the end of the article, other than to say an SA founder did something like this (he didn’t), when you clearly acknowledged that the propulsion is done via external linear induction motors?

    Very sloppy article. Is the editorial process just entirely shelved for blogs? I would be absolutely ashamed to have my name attached to this.

    Link to this
  11. 11. ScienceIsWinning 3:20 pm 08/15/2013

    As for the cost of the land: I’m not sure how the vast bulk of the land required being on the already public I-5 and I-580 with the remaining land required for purchase not being entire swaths of land but only being the footprint of the pylons was lost on you. The only explanation I can see is that you didn’t bother to look over sec 4.4 either.

    Link to this
  12. 12. shazam 6:06 pm 08/15/2013

    My claustrophobia is kicking up. Squeezed into a pod, roaring along in a tube, with G forces in all directions and no outside reference…Where’s the barf bag?

    Link to this
  13. 13. Lun_Esex 8:42 pm 08/15/2013

    Here’s a good read for detailed, technical, and practical info and criticism about Hyperloop:


    “pylons are expensive, as can be readily seen by the costs of elevated highways and trains all over the world. The unit costs for viaducts on California HSR, without overhead and management fees, are already several times as high as Musk’s [500 km of viaduct for $2.5 billion] cost …”

    “[Hyperloop] proposes a lateral acceleration of 4.9 m/s^2: 0.5 g. This is after canting … 0.5 g, or 4.9 m/s^2, is extreme. Non-tilting trains do not accelerate laterally at more than 1.2 m/s^2 in the plane of the track (i.e. after accounting for cant), and at high speed they have lower lateral acceleration, about 0.67 m/s^2 …”

    “At 1,220 km/h, it is very hard to climb grades, which would require very tall viaducts and deep tunnels under mountains. Climbing grades is easy, but vertical acceleration is such that the vertical curve radius has to be very large. A lateral acceleration of 0.67 m/s^2 would impose a minimum vertical curve radius of 170 km, versus 15 km at 360 km/h HSR speed. …”

    “If vehicles brake at a constant rate, the safe headway is half the total deceleration time; if a vehicle brakes from 1,220 km/h to zero in 60 seconds, the average acceleration is more than 5 m/s^2, twice the current regulatory safety limit for passengers with seat belts. …”

    “If Musk really found a way to build viaducts for $5 million per kilometer, this is a huge thing for civil engineering in general and he should announce this in the most general context of urban transportation, rather than the niche of intercity transportation. …”

    There is much more at the above link, and it comes with citations, which is something Elon Musk’s Hyperloop proposal appears to be lacking.

    Link to this
  14. 14. pragmatickid 10:47 pm 08/15/2013

    I LOVE IT. but nothing new about this idea. (MUCH BETTER GRAPHICS, BUT impractical).
    60, yes 60, years ago, (I am 85+) I developed, as a young engineer, at Cornell Aeronautical Lab, a delta P transit system, based on the OLD vacuum transit tube concept universally used to send communications around a building. It was in common use at the NY Public Library, the Pentagon and many department stores. It’s fantastically simple and there is no excessive heating, The cars were inexpensive, unpowered and safe. this was because if a car closed in on another, the pressure between them would rise, slowing the rear car. It would take far to much time to go into all the details, here, we built a 300′ model tube in which we achieved 120 f.p.s. equivalent to a full scale velocity of about 600 f.p.s. All aspects were covered including passenger comfort and safety. Stopping and starting were achieved by gates at the stations. Air compressors at the stations supplied the propulsion air augmented by the air recovered by the train stopping at the gate. It was much fun to work on, but no one was really interested in it at the time.
    Another great idea was a catenary tube linking two cities. However it required a the tube to go 20 miles into the earth to go from N.Y. to D.C. but it could, theoretically, do it in ~25 minutes with very little added power; just enough to overcome the aero friction. Sort of imagine Tarzan swinging through the Earth with the cities replacing the trees.
    OH, MY.

    Link to this
  15. 15. MCollins 12:18 am 08/16/2013

    In the above linked blog, Pedestrian Observations ( one of the commenters makes the following priceless post…

    “Keep Houston Houston says:
    2013/08/14 at 16:13

    Hyperloop doesn’t sound like fun. So I cram into this pod that’s 4.3′ wide – two feet narrower than my Crown Vic. Then I recline way the fuck back because the ceiling is too low to stand up. Of course this means bathrooms are a foregone conclusion. I look forward, but you can’t see in front of you. I look out the side, but the tubes are solid metal – yeah, they were glass in the visuals, but that shit got value engineered out right quick. Whatever. Nothing to see. Only going to be 30 minutes until we’re in SF, right?

    I reach over to grab my girl’s ass, except I can’t, because a big chunk of the propulsion is in between the seats. So I fold my hands across my chest to enjoy the g forces which I’ll spend the next 30 minutes experiencing with no visual frame of reference. Thankfully, I haven’t eaten in several hours. But the guy sitting behind me thought it’d be good to get a double double before he got on the train, and after 10 minutes of being thrown around the inside of the pod, he pukes.

    Of course he can’t take the barf bag to the bathroom because there is no bathroom. I look at my female companion and scratch her shoulder since that’s about all I can reach. Over the next 20 minutes two more people barf as the cabin fills with the smell of upchuck. We arrive in SF. She suggests maybe next time we just take my Ford up the Grapevine. I emphatically agree.”

    Link to this
  16. 16. Tandem78 11:15 am 08/16/2013

    I am not happy with the idea of having nearly 300Kg of compressed steam in the vehicle. What happens if the pod has to stop in the tunnel in an emergency. There is no effective way of cooling the vehicle down, and at some point the steam will go into thermal equilibrium with the rest of the pod.

    Link to this
  17. 17. psngray 11:41 am 08/16/2013

    I love this idea, there are bugs as they’re are with all concepts in the alpha stage, but basically the concept is sound. Would people put up with being sealed in a tube for 30 minutes with no windows? They might for a fare of only $20. Windows could be replaced with screens, a 4k screen is very realistic.

    I think it’s worth building a test track to work out some of the bugs, and get more familiar with the hardware and problems.

    Link to this
  18. 18. bucketofsquid 1:54 pm 08/16/2013

    OK, so these vehicles travel on a cushion of air like a hovercraft. They are propelled by electromechanical boosters and travel in a low pressure tunnel. How does this compare to airline travel or a basic maglev train?

    Since regular rail can move things as much lower cost than a semi, why not just expand the rail net and develop car carriers? That way you can go between cities and drive around once you get there.

    Link to this
  19. 19. pragmatickid 3:44 pm 08/16/2013

    Please read comment 14. In answer to some of the other stated objections, the CAL loop used NO steam or other onboard propulsion, at full scale it would have been 8′ in diameter, acceleration forces and jerk (Rate of change of acceleration) would have been about the same as a modern sports car, and safety would exceed any current transportation mode. The only drawback was the claustrophobic effects of a closed, windowless room; similar to riding an elevator for an hour.

    Link to this
  20. 20. pragmatickid 3:59 pm 08/16/2013

    I thought I might add that I was the program manager for the Navy’s
    Surface Effect Ship (the World’s Fastest Warship (see Guinness Book of World Records) program. It was a hard sidewall cushion craft that exceeded 90 knots. It needed the sidewalls so it could have more efficient water propulsion, unlike true hovercraft, that require air propulsion. Working in this area pretty much proved the difficulties of a tubed or channeled hovercraft. As far as Maglev, since it is nothing more than a straightened out motor, it’s costs are incredibly high, but within range for an aircraft catapult, which the Navy is now developing.

    Link to this
  21. 21. Heteromeles 1:56 pm 08/17/2013

    Oooh boy, start the giggling.

    To start with, I-5 is (or used to be) an autobahn in the central valley. That was the only place my friends and I ever tried to max out the speed on our cars. We went over 100 mph. Yes, there are some cops with very fast cars, and yes, we were lucky not to lose control and die, but it’s flat out there. On the same trip, we took a ruler along just to see if it was perfectly flat (it was close).

    That said, there’s a little obstacle between LA and the Central Valley called the Grapevine, where the road goes from 1600 feet in the valley to 4183 feet at Tejon Pass 12 miles later (this stretch is notorious for brush-fires sparked by burning car engines and brakes). It then wends its curvy way through the mountains before dropping back to around 2000 feet in the LA Basin.

    In the proposed hyperloop, this would be…exciting at Mach one. I don’t think even fighter jets try it, at least not routinely.

    Then there’s safety. Mr. Musk wants the cars five miles apart, and if one crashes (say, because the tube fails), the cars behind it have to decelerate at over 2 G’s laterally to avoid crashing. That’s eye-popping, perhaps literally. I don’t ever want to be hanging from my straps decelerating at 2-Gs face forward.

    Of course, we’d also have to figure out how to extract people from cars stuck in the tube, which might mean depressurizing the whole system and bringing traffic to a standstill throughout, probably for days until it could be repaired and repressurized again. Otherwise, cars just keep decelerating into the mess until their weight overwhelms the monorail supports, and the tube buckles to the ground. This is the mass transit version of what naturally happens in the Valley during Tule fogs, when massive multi-car pileups are common.

    Then, of course, there’s the problem of maintaining that vacuum inside the tube. Here, I’m thinking of terrorism. G. Gordon Liddy back in the 1980s pointed out how few people it would take for him to take down LA, by dynamiting I-5 overpasses in the Grapevine, sending a raft of dynamite through the aqueduct tunnel to blow up inside the mountain, and cutting a few critical power transmission towers in the same area. Did I mention the Grapevine is remote, far from large towns, and extremely rugged?

    So let’s look at what someone with a hunting rifle would do to the hyperloop (or a Barrett 50 cal, if you want something that’s more illegal in California). Right, perforations are bad for the hyperloop. It has to be bulletproof, and it also has to be monitored to an enormous degree, because some dufus might just plant dynamite on a couple of critical pylons and blow them up right before a capsule comes through.

    Oh yeah, and don’t forget that this crosses the San Andreas fault. Possibly twice, depending on where it comes to rest near San Francisco. Even with out major quakes, the faults creep at a fairly regular rate, so the tube would have to be rebuilt every year or two where it crosses, or it will warp and fail. I’m not sure what the tolerances are for something this extreme, but if they’re very tight, they might have to rebuild the hyperloop over each fault every year, just to accommodate the shifts along the fault.

    So all I have to say is: brilliant, Mr. Musk. Might I suggest building it between, oh, Cincinnati and Toronto via Detroit? It’s much flatter there, much less geologically active, and such a mammoth construction might, oh, I don’t know, revitalize that part of the Rust Belt. Just saying.

    Link to this
  22. 22. Satya Narayan Tiwary 3:07 am 08/22/2013

    From Bullock cart to Bullet train to Hyperloop transportation systems, i.e., past, present and near future systems, journey, in travel world, seems to be exotic development. In far future, super Hyperloop, faster than Hyperloop, may be developed. Thus, speed is increasing. Speed limit is speed of light. Can we move with speed of light? Think ahead of time. Be ahead.
    S. N. Tiwary, Director
    Former Head, Dean, V. C. (acting)

    Link to this
  23. 23. HarryKeller 1:42 pm 08/30/2013

    This is certainly a “visionary” idea. It may recycle some very old ideas but buttressed with new technology and with some new wrinkles. Every single technological issue has not been thoroughly worked out, but a great many have.

    MCollins quotes someone who echoes what I wrote on ETCJournal (.com). The primary issue I see with this system is being strapped into a tiny space with almost no wriggle room for 30 minutes. Without any view, it will quickly become claustrophobic. However, for a small percentage of the overall cost, you can fix that problem readily.

    What I’d like to see is someone with Musk’s vision taking on the issues of science education in the United States (and the world too). NGSS doesn’t do it, but then I didn’t expect a committee to hit a home run. This is my area, where I’m working now. With Musk’s resources and a few more visionaries, we could truly transform learning science. That would automatically (if done properly) fix reading, writing, and math skill issues. There is overlap with social studies, music, art, and physical education.

    Forget No Child Left Behind. Welcome to every child inquires, explores, and discovers.

    Link to this
  24. 24. bsebaloaf 5:34 pm 09/16/2013

    Vertical Hyperloop / Space elevator
    Has anyone considered using the Hyperloop idea over at SpaceX? It’s only 60 miles up to space, and they’re trying to find cheaper ways to get payload off the ground, how about building that loop vertical rather than horizontal? Or on a ramp. It wouldn’t have to be any longer than the proposed loop from LA to San Fran, could even be shorter.

    Link to this
  25. 25. BMGalvarado28 7:27 pm 09/26/2013

    The Hyperloop looks to be an exciting option for fast distance traveling, but similar to current public transportation options like San Jose’s light rail system or Caltrain, it relies on the public to travel to and from specific destinations. Our BiModal Glideway would be a better solution as it would provide the same high speed distance travel, but allow drivers the freedom to go from door to door during their commute.

    Link to this

Add a Comment
You must sign in or register as a member to submit a comment.

More from Scientific American

Email this Article