High-Speed Rail or Hyperloop? Let’s Try Both, and Reward the Winner

High-Speed Rail or Hyperloop? Let’s Try Both, and Reward the Winner

It’s been a bumpy week for high-speed rail advocates in California. On Sunday, the Los Angeles Times published a story saying that groundbreaking on the first section of the “bullet train” route linking L.A. and San Francisco will likely slip into 2014—more than two years behind schedule. Then on Monday, aerospace and automotive mogul Elon Musk published details of his proposed Hyperloop, which would shorten the one-way trip from L.A. to San Francisco to just 35 minutes by shooting passenger pods through steel tubes on cushions of air at more than 700 miles per hour.

No one thinks the timing was a coincidence. It seems clear that Musk is trying to deflate enthusiasm for the existing rail project by dangling an even sexier alternative in front of the traveling public’s eyes. “I don’t think we should do the high-speed rail thing,” Musk told the San Francisco Chronicle. “It’s basically going to be California’s Amtrak.” (“He didn’t mean that as a compliment,” Bloomberg Businessweek helpfully explained.)

But it’s a false choice. The proposed routes for the two systems don’t overlap by much, so building one doesn’t preclude building the other. It’s really a question of hedging our bets—and supporting innovation of all stripes.

The 200-mph rail link between Los Angeles and San Francisco, approved by California voters in a 2008 ballot measure, will cost at least $68 billion overall and will start operating over its full length by 2029 at the earliest. Its main advantage is that it uses proven technology: after 180 years, we know a lot about building railroads. The Hyperloop concept hasn’t advanced much beyond the sketch-on-a-napkin stage, so it would need to be thoroughly tested. But if, as Musk says, his proposed system can be built faster and for far less cash—$7.5 billion, which is little more than a rounding error in the California High Speed Rail Authority’s project—then there’s no reason to choose between the two ideas. Logic and economics dictate that we should try building both, and let the chips fall where they may. If we find out after a few years that Musk is right, we’ll still have time to pull the plug on high-speed rail. At that point, no one will want to waste $68 billion on an unneeded rail line that will take another 10 years to complete.

In fact, to up the stakes, I’d like to propose a prize competition, of the same sort that spurred John Harrison to solve the longitude problem in the 18th century and Charles Lindbergh to complete the first solo transatlantic flight in the 20th. Here’s my idea: the first organization to deliver a live human from Los Angeles to San Francisco over a fixed ground route in under three hours wins $10 billion. We can call it the Smog to Fog Challenge.

Okay, the name might need some work. Also, I’ll need to find a group of government agencies, corporations, and philanthropists enlightened enough to stake the prize money. (Larry and Sergey, I’m looking at you.)

While the title and the prize amount I’m suggesting are a bit tongue-in-cheek, I’m not joking about the idea of a prize. NASA, DARPA, and the X Prize Foundation have shown over and over that prize-based competitions stimulate innovation. And competitions have a venerable history in the railroad business. The 1862 Pacific Railroad Act, which authorized the construction of the Transcontinental Railroad, rewarded the competing companies—the Union Pacific and the Central Pacific—with 12,800 acres of land and $48,000 in government bonds for every mile of track they laid toward the central meeting point (which turned out to be Promontory Summit near Salt Lake City, Utah). Union Pacific, which had the easier job, won handily, completing some 1,086 track-miles from the Missouri River westward to Utah. But Central Pacific built an impressive 690 miles of track stretching from Sacramento eastward over the Rockies to Salt Lake.

It was, arguably, the best money the federal government ever spent. There’s no need to detail how much economic value was unlocked by the completion of the transcontinental link—the first in what became a vast network. Sending passengers and freight across the continent on iron rails was such a good idea that we’re still doing it today, long after nearly every other 1830s-era innovation has become obsolete.

Rail travel is fast, clean, efficient, and still a little romantic. That’s why I was a model railroader as a kid, and it’s why I’m a longtime fan of the California high-speed rail project, whose trains would glide down tracks just one block east of my loft building here in San Francisco. You have to admit that the idea of being able to travel from San Francisco to Los Angeles by train in just two hours and 40 minutes—compared to an interminable five and a half hours by car—is pretty attractive. So is the project’s carbon profile. The High Speed Rail Authority has said that the trains will run wholly on electricity from renewable sources. And it has projected that by keeping people out of their cars, the rail project will lower greenhouse gas emissions by some 4 to 8 million metric tons every year—the equivalent of shutting down a coal-fired electric plant.

Artist's rendition of the California high-speed rail line on its approach to Bakersfield, CA.

Artist’s rendition of the California high-speed rail line on its approach to Bakersfield, CA.

Granted, the high-speed rail plan is far from perfect. The route deemed most politically viable is also frustratingly indirect, taking LA-to-San Francisco passengers on a wide detour through Bakersfield and Fresno. (Phase 1 of the project, in fact, is a 300-mile “backbone” section connecting Merced to the San Fernando Valley, meaning nobody in San Francisco or Los Angeles will actually see high-speed trains until the late 2020s.) And for passengers departing San Francisco after 2029, the first 50 miles of the trip will hardly feel bullet-like. The authority has decided that it would be impractical to build new high-speed tracks in the existing Caltrain right-of-way, meaning the new trains will have to trundle down the peninsula at commuter-rail speeds.

But none of that is the stuff that bothers Musk, the charismatic co-founder of PayPal and the CEO of both Tesla Motors and SpaceX. His principal objection is that we already have a fast way to get from San Francisco to Los Angeles: flying. Unless a new transportation system is faster, cheaper, and safer than air travel, Musk argues in a 57-page design brief posted this week on the Tesla Motors blog, it’s not worth building. (He adds that a new system should be more convenient than flying, immune to weather, sustainably self-powering, immune to earthquakes, and “not disruptive to those along the route.”)

The big question, of course, is whether the Hyperloop would qualify on any of these counts.

The system’s only indisputable advantage, based on the designs Musk has offered, is that it would be faster than plane travel, and way faster than train travel. Here’s the basic idea. A Hyperloop “capsule” would be narrow—only about 4.5 feet wide and 6 feet high—and would carry 28 passengers in a single row of seats. At stations, the capsule would roll on wheels, but in transit the wheels would retract and the whole assembly would float on skis that emit a thin layer of pressurized air.

The air would come from the Hyperloop tube itself. To reduce drag on the capsules, pumps would lower the air pressure inside the tubes to 100 pascals—about the same as the pressure at an altitude of 150,000 feet. But there would still be enough air in the tubes to be scooped up by a battery-powered compressor at the nose of the capsule and sent to a pressurized storage tank, and from there to the skis.

Some of the air would also spew out the back, providing a bit of thrust, but the capsule’s real propulsion would come from a linear induction motor system. The design calls for “rotors” or aluminum fins on the outside of the capsule to provide acceleration as they pass through electrified stators attached to the inside of the tube.

The power for the stators and air pumps would come from solar panels installed along the top of the pair of tubes (northbound and southbound). The tubes themselves would be placed 20 feet above the ground on concrete pylons spaced 100 feet apart. The Hyperloop’s main footprint on the landscape, then, would be a chain of 25,000 pillars from Los Angeles to San Francisco, mostly along the median of I-5.

Musk says the capsules could travel at 300 mph along I-5’s curvy sections and a near-transonic 760 mph along the straightaways, for a total L.A.-to-San Francisco travel time of 35 minutes. Well, technically, Musk’s suggested route for the Hyperloop stops in Oakland—there is no provision for getting the capsules across San Francisco Bay. So in an apples-to-apples comparison between the Hyperloop and the high-speed rail plan, you’d need to add another 15 minutes or so for the decidedly subsonic subway ride from Oakland into San Francisco.

Conceptual design sketch of Hyperloop passenger capsule.

Conceptual design sketch of Hyperloop passenger capsule.

If anybody has offered a convincing case for the Hyperloop’s technical infeasibility, I haven’t seen it yet. Nothing Musk proposes violates the laws of physics or requires brand-new technology. It’s hard to know in advance how much energy would be required to pump air out of the Hyperloop tubes and to compress it onboard the capsules, so it may be optimistic to assume that batteries and solar panels could supply all of it. On top of that, some critics say drag and friction inside the tubes might be unmanageable, while others worry that passengers would be exposed to excessive g-forces. But such details could probably be ironed out.

So the next big question is cost. Musk says a fleet of 40 capsules could be built for $54 million, and the tube system for $4.4 billion. He figures the land required, apart from the I-5 median, could be bought for another $1 billion, for a total project cost of only $6 billion. A larger version of the Hyperloop capsule and tube system, capable of carrying vehicles or freight, specs out at $7.5 billion. If you amortized that cost over 20 years, the price of a one-way ticket would be only $20, Musk calculates.

Now, I’m neither an economist nor an engineer, but my instinct as a technology observer is that Musk has probably low-balled all of these cost estimates. Just look at the AirTrain system in New York City, an elevated railway system that carries passengers from from a subway station in Queens to JFK Airport. Like the proposed Hyperloop, the AirTrain runs along a viaduct above a highway (the Van Wyck Expressway); also like the Hyperloop, it uses linear induction motors for propulsion, and is fully automated, with no engineers or conductors on board. That system, built by the Canadian rail equipment firm Bombardier Transportation, cost $1.9 billion, and it’s only 8 miles long. At New York prices, covering the 350 miles between Los Angeles and San Francisco would cost an eye-popping $82 billion—about $14 billion more than the estimated cost of California’s high-speed rail project.

If Musk’s estimates are off by an order of magnitude, which doesn’t seem unreasonable, the Hyperloop will cost more than high-speed rail (assuming that the rail project itself doesn’t continue to explode in cost). If he’s only off by a factor of five, the Hyperloop will be cheaper. Either way, it’s too early to say whether the Hyperloop is or isn’t an economically viable alternative to the rail project. The only way to find out is to try building it, starting with extensive computer simulations, followed by the construction of a test track.

This will involve no small risk and expense, given the distances required to accelerate an 8,400-lb capsule to 760 mph, then slow it back down. Even a tycoon like Musk can’t finance something like that out-of-pocket. California voters set aside $9 billion in state bonds for the rail project; to level the playing field and jumpstart the market for a next-generation land-based mass transportation system, we need an incentive prize like the Smog to Fog Challenge.

If the California High Speed Rail Authority finishes first, great—winning the prize would reduce the cost to taxpayers by $10 billion. If some hypothetical Hyperloop Consortium wins (I’m seeing a collaboration between Tesla, Bombardier, and Lockheed Martin in my crystal ball) then we end up with a really cool way for hipster documentary-makers in San Francisco to zip down to Hollywood for the Academy Awards or for Angelenos to grab dinner at the Slanted Door.

But who could fund such an enormous prize? I’d like to nudge the big Los Angeles and Silicon Valley biotech, defense, Internet, and computing companies, who have a vested interest in the quality of life in California, to step up. For if nothing else, a system as fast as the Hyperloop could be seen as a really fat data pipe. Musk thinks a Hyperloop capsule will be able to carry passengers and luggage weighing up to 2,800 kilograms. If you replaced all of that weight with 1-terabyte hard drives and rushed them from L.A. to San Francisco in 35 minutes, you’d get an effective bandwidth of 4.8 terabits per second. That’s about 50 times faster than today’s state-of-the-art 100-gigabit Ethernet networks.

So that’s my proposal. I went into my study of Musk’s design document thinking that the Hyperloop was, well, loopy, and that the hordes of instant critics this week were probably right. I came out thinking that the concept might just be feasible—and hoping that Californians, and Americans, haven’t lost the ability to think big. Let’s keep moving forward on high-speed rail—it’s one of the only audacious projects on the docket right now. But let’s also give Musk some room to put the Hyperloop to the test.

The Author

Wade Roush is a contributing editor at Xconomy.

By posting a comment, you agree to our terms and conditions.

  • DocScience

    Let’s not get confused about WHERE high speed rail can go and WHERE hyperloop can go.

    HSR is a multi-hundred ton behemoth with ugly overhead electric lines and towers, creating incredible noise, danger and requiring immensely strong, large and expensive rail mounting structures, like a huge bridge to be above ground or a hundred foot wide path to be on-grade.

    Hyperloop is two self-supporting and relatively cheap 7 foot tubes supported every 300 feet with a small post, silent, with a lightweight pod which doesn’t have the energy to escape the tube in a “wreck”.

    Hyperloop can go places HSR can not.

  • jan

    Why 10B? Would not 2B be enough? It does not have to pay the full cost. It should just spur the innovation, competition and lure private investors.

    The price as proposed can be read like this: one can build something for 7B today and cash 10B tomorrow. Good investment. Done. Who care about future?

    At 10B there is no incentive for profitable operations if the price pay 100% investment. It’s just a rush to build _something_ first. It can break a week later. It can be unreliable or even unsafe.

  • DisquisTL

    I’m pretty sure that I’m one of the majority of Californians who has voted down the “high speed rail” line: idea, study funding, and funding. In fact, we voted this sucking hole down three times now.

    At each stage, the idiots behind the idea have continued to convince legislators to dump public funding into the thing, despite being rather loudly told by the public Do Not Fund This Thing We Do Not Want And Will Not Use.

    I live in San Francisco. If I want to go to LA, I’m not going to ride this “high speed rail” that isn’t actually high speed, starts a long bus ride away from where everyone lives, and goes some place no one wants to go (Bakersfield; seriously?).

    Every time someone brings up fast transportation between two points, someone else insists on bringing up something totally and completely unrelated to fast transportation between two points: the California High Speed Rail Authority.

    If it’s such a great idea, then why does it have to be publicly funded? People thought the new Bay Bridge was a good idea, and look how that’s turned out. No thank you on the “high speed rail”.

    When it comes to the hyperloop: the primary advocate isn’t a politician, he’s an engineer. In fact, he’s not just an engineer, he’s a freaking rocket scientist. In fact, he’s built some majorly successful engineering companies. In fact, he’s built the only private company in the history of mankind to launch something into orbit. In fact, he’s built the only private company capable of launching resupply missions to the ISS: something that right now, NASA can’t do.

    So forgive me a bit for trusting his judgement here.

  • http://www.xconomy.com/san-francisco Wade Roush

    Thanks, all, for your comments so far. What’s interesting to me is how much venom Musk’s proposal is stirring up on both sides: from people who absolutely detest the high-speed rail project, and from people who are incredibly angry at Musk for trying to derail it (so to speak).

    Jan: I’m note sure that $10B is the right amount for a prize; maybe $2B is closer to the mark. (All amounts are theoretical in the absence of a donor.) But whatever the exact amount, it should be big enough to work as a real monetary incentive, not just a pat on the head.

    One reader sent in a link to a blog post that contains some pretty interesting initial criticisms of the Hyperloop. I particularly suggest scrolling down to the comments from Clem Tillier: http://www.cahsrblog.com/2013/08/hyping-the-hyperloop/

  • thewholesickcrew

    One of the weaknesses in CA HSR is whether or not it can get the ridership required to make it self sustaining. The cost and transit times are two big factors on whether or not ridership will be there. Initially in 2008 they sold CA voters on $50 tickets (from LA to SF) on a train that would reach 220 mph and make the trip in 2 hours and 40 minutes. However, today all those claims are in serious doubt. The costs are rising to build it thus rising projected ticket costs. There are serious concerns that the train will not be able to reach 220 mph thus not making the trip in 2 hours and 40 minutes. The proposed blended rail system put forth by Jerry Brown will only slow HSR down by lowering costs by having it share lines with slower commuter/freight trains. This all leads to a system that is not competitive with flying. At the end of the day HSR may not be cheaper or faster than flying.

    Now if Musk is correct and hyperloop can have transit times of 30 minutes between LA and SF I believe he would have a winner. At that transit time I believe it would garner the type of ridership needed to have it support itself and not become California’s Amtrak. At commute times of 30 minutes that would open up a whole new concept of commuting to work. I would like to see more details though on hyperloop before being completely sold. It is true we know how to build HSR and we see examples around the world. Hyperloop has not been proven. I would like to see a working prototype operate before being sold on it.

  • Joshua Zev Levin, Ph.D.

    A plague on both your houses!

    Both proposals limit you as to where you can go at high speed. If you don’t live right across the street from both of the stations, then you have to spend additional time, and your HSR is not so high-speed anymore. You may have to travel dozens of miles.

    The LeviCar proposal (www.LeviCar.com) means that you drive not more than about 10 miles (in built-up areas, less in congested areas, more in wilderness areas) to a depot. Then, you can ride in your own car body at 300 mph, on a superconducting MagLev (Magnetic Levitation) network, to another depot within about ten miles of your destination, and drive from the depot to there. For freight, the RoboTrail service would use the same network but different depots. Resorts, hotels, and warehouses could be directly connected to MagLev lines.

    Simple calculations will show that LeviCar and RoboTrail can give faster door-to-door service, for most customers, than airlines, HSR, and Hyperloop. LeviCar and RoboTrail are OD-HSR (On-Demand High-Speed Rail.)

    Before replying to this pose, please check out http://www.LeviCar.com for more details.

  • Colin Tree

    In all cities we need different stages of public transport starting with small, flexible suburban transport from the residence or business, linking with inter suburban medium volume and speed, linking with inter metropolis high volume, high speed.

    The interchanges between each stage is an important factor for efficiency.

    Computerised booking and flexible timetables also add to efficiency. We don’t need fixed timetables if we can make a good demand driven system.

    No need to go very fast if you have to wait 1 hour to ‘fly’, which is currently the case at all airports.

    More importantly is using communications to save the need to travel and decentralisation to reduce bottlenecks.

  • http://www.kipling.org.uk/poems_macdonough.htm bobw

    Unfortunately, passenger rail is *not* efficient, unless you can get the ridership up. Historically only commuter lines have been profitable. The railroads only offered passenger service because it was expected, when it didn’t lose money too badly. When it started losing money even faster they discontinued passenger service. Hence Amtrak.
    Profit is not evil. It is confirmation from external reality that your output is worth more than the sum of your inputs. You are not merely rearranging deck chairs on the Titanic.