Elon Musk seems to be a combination of a Horatio Alger hero with Tom Swift — a self-made billionaire with PayPal, he now is devoting his time and energy and his own money to projects like SpaceX and Tesla Motors that sound like the science fiction of the ’50s. Electric cars? Commercial rockets? Capitalism?
Rockets that take off and land vertically?
His newest idea, released after some weeks of speculation, is a super-high-speed train concept he calls Hyperloop. The excitement lasted for almost a day before critics started to decry the idea. Alexis Madrigal at the Atlantic manages to sum up both attitudes in two paragraphs:
So, two thoughts on the Hyperloop, which I find to be in tension. First, like anyone who has ever read a sci-fi novel or made the sound “pew-pew” with a raygun made from thumb and forefinger, I think is fantastically cool and wonderful. A pod system that shoots you to LA! Amazing! Even the drawings evoke that ’50s can-do futurism. There’s none of that dark ’60s/’70s technoanxiety in this proposal. None.
He’s right here: Musk doesn’t appear to believe in being limited.
Which brings me to thought two: I worry that more fully baked transportation projects might be put on hold in hopes that Musk’s still-fictional idea works out. Musk’s proposal, because of who Musk is, could serve as a poison dart for California’s high-speed rail, and then nothing comes of it, leaving the state with an outdated passenger rail network and no Hyperloop to make up for it.
His point here being that he’s afraid that the existence of a proposal for a financially sound, technically feasible supersonic train will hinder the development of a financially unsound, 130 m.p.h. train system that will cost ten times as much.
Of such worries are modern “futurist” careers made.
That said, the Hyperloop idea does have some problems. So let’s look at the idea itself first, and then look at some pro and con arguments.
The Hyperloop approach is simply the “tube trains” of science fiction going back to Hugo Gernsback’s Ralph 124c41 — people in pneumatic tubes, traveling at high speed. The Rand Institute looked at a notion like this, using tunnels at very low air pressures, forty-odd years ago. (The Rand proposal fell down on one point: digging tunnels is expensive.)
Musk has put engineering minds to work on how such a concept could be made both technically and financially feasible. The first thing to go was tunnels (except possibly for comparatively short tunnels to cross physical obstacles). Tunneling is expensive — I know, I said that, but tunneling is expensive, in the neighborhood of $100 million dollars per mile — so tunnels had to go. Instead, the Hyperloop “tunnels” would be steel pipes on pylons mounted well above ground — which simplifies another problem, because instead of needing to buy land on which to run the trains, the Hyperloop can simply follow existing highway rights of way, at least most of the way.
Inside the tunnels would be, well, nothing. Or darn close to nothing. The air pressure would be reduced to 100 Pascals — 0.01 pounds per square inch, or about one one-thousandth of normal atmospheric pressure. Musk’s note on Hyperloop notes this isn’t a real vacuum, and to a space guy it’s not — but it’s one tenth of the atmospheric pressure at 100,000 feet altitude, or one fifth of the atmospheric pressure on the surface of Mars. To us civilians, that’s a pretty hard vacuum.
The trains themselves are single cars that would look a good bit like the body of a private jet — streamlined aluminum tubes that fit inside the steel tunnel of the Hyperloop. At the speeds Musk envisions, up to 800 miles per hour, even the tiny air pressure inside the tubes makes aerodynamics an issue, so this becomes a design constraint: shock waves could build up between the cars and the inside of the tubes and produce a tremendous increase in drag. So the tubes must be big enough that the car is no more than about six tenths of the inner diameter of the pipes — a seven foot diameter car needs an eleven foot tube.
At 800 miles per hour, wheels are really Right Out, but there are other options. Musk’s engineers chose an air cushion: compressed air flows out of special skis against the walls of the pipe, lifting the car away from the surface. (I have some other ideas; more on that below.)
Motive power comes from the tube itself. Linear induction motors — basically a normal electric motor, but long and flat instead of round — accelerate the cars up to speed and brake them at the end. Using this scheme has a lot of advantages, one of the biggest being that the braking actually puts energy back into the system. Our unwound motor becomes an unwound generator, so much of the electricity that went into getting the car moving is recovered in slowing it down.
Where does the electricity come from? For once, we see a good use of solar energy. The tubes on their pylons are a good place to put solar panels; the total length of the route means that literal square miles of solar cells could be put on top of the tubes. There would need to be batteries to store power for surge loads and against rainy days, but Musk’s group did the calculations, and the capacity demand could be handled with commercial batteries available today. (Like the batteries in the Tesla, in fact.)
Although Musk doesn’t mention it, the solar energy might even help pay for the system. If you expand the solar cell arrays to be 100 feet wide, my back of the envelope estimate says you could collect in the neighborhood of 3 gigawatts; use what you need and sell the rest.
All the technical ideas make sense, but I think they’re missing a possibility or being cagey about it. The air-cushion skis mean that each train is constantly pumping air into the tubes, which then has to be pumped out constantly. If the compressor that is providing the air fails, the car drops on its skis against the inside of the tube.
At 800 miles per hour.
This will, ahem, strike some sparks.
There is another option, however. Scientists and engineers at Stanford, University of Pittsburgh, Lawrence Livermore Labs, and other places have a concept called Inductrack, where permanent magnets on a train car interact with unpowered passive conductors in the tracks to magnetically levitate the train car. Other maglev trains have been excessively expensive and required a lot of power; Inductrack avoids those problems. The down side is that the car has to get up to a certain minimum speed before the levitation lifts the car off the track, but that minimum speed is very low. It does mean the cars need “landing gear” on which to roll for the first and last few few hundred feet.
Elon Musk, call me; I’ll be happy to make a deal on this idea.
The technical feasibility of it, at least at the level of Musk’s 57-page report, seems clear. The problems seem primarily, well, political.
For reasons that frankly remain unclear to me, there has been a lot of devotion to the idea of high-speed train systems so that the U.S. would have a train system as good as Europe. My best guess is that people in the Ruling Class tend to stay close to the Boston-Washington corridor, and forget that all of Europe is only about three times the area of Alaska. As I wrote back in 2008, even high-speed trains just can’t compare with airlines over the area of the U.S. (What’s more, airlines are beginning to take over long trips in Europe too — even the French LGV trains take four times as long to travel the same distance.)
Trains are also extremely expensive to build: Musk estimates the cost of his San Francisco-Los Angeles Hyperloop at $6-$10 billion; the proposed high-speed rail in California is ten times as much for five times longer travel times. The Reason Foundation published a study last April showing that the system as it’s currently proposed would lose between $124 million and $373 million a year. What’s more, according to the study:
This report updates Reason’s 2008 Due Diligence Report by addressing and evaluating numerous changes in California’s plan to build a high speed rail (HSR) system between San Francisco and Los Angeles via the San Joaquin Valley. This Due Diligence Update addresses the Authority’s revised documentation, business plans and public statements issued between 2008 and late-2012, which are found to be similarly inaccurate, misleading and in violation of the laws guiding the project. Additional analysis is warranted to respond to the Authority’s newer yet illusory capital cost reductions, likely capital cost escalations, need for operating subsidies, slower train schedules, high ridership projections, and the inability to meet the statutory requirement to link Los Angeles and San Francisco in 2 hours and 40 minutes or less.
A system at Hyperloop prices can pay for itself and make a profit; a system like the conventional high-speed train never will.
But there is clearly a substantial constituency for conventional high-speed trains, primarily a political one. During the election campaign for the California state bond issue to finance the initial construction, 97 percent of the funding for Californians for High-Speed Trains came from the 40 top contributors; of those 40 top contributors, 39 of them were construction companies, construction unions, train equipment companies … and California’s electric utilities companies, which would be expected to provide electricity for the all-electric train system.
There are other political issues as well. Musk’s plan would use the existing highway right of way, putting the elevated tubes on pylons above Interstate 5 — but those rights of way are controlled by the state of California, through CALTRANS. The “high-speed” rail system has been a centerpiece of Governor Jerry Brown’s campaign, and has been very firmly and publicly supported by President Obama and Vice President Biden.
It seems possible that CALTRANS might not be very anxious to provide Hyperloop with access to those rights of way.
Maybe I’m just excessively cynical, but it seems to me that with a choice between Musk’s really high-speed train system, and a “high-speed” train system crippled by design compromises that would funnel $68 billion (at last estimate) through the unions and companies that lobbied for its construction, it might not be Musk’s system that would win politically.