The hyperloop: if you think the Delmar Loop Trolley in St. Louis, MO was a boondoggle, just wait till the state starts hemorrhaging cash for the proposed hyperloop test track, and later a possible route connecting St. Louis, Columbia, and Kansas City. The hyperloop would rely on magnetic levitation (maglev) technology that has been used for trains in some parts of the world, though always on relatively short routes. For a hyperloop, however, the maglev system keeps carrier “pods” suspended in a near-vacuum tube extending the length of the route, eliminating friction and air resistance. Proponents say the pods will move at top speeds of 700 miles an hour, traversing the state in about 30 minutes. And they say it will be a very green machine.

Richard Branson’s Virgin Hyperloop One wants to build the 15-mile test track, which is projected to cost $300 – $500 million. That range is centered just a bit higher than the cost of the Loop Trolley on a per-mile basis, and for a project with major technological uncertainties, that leaves me just a bit wary. The 250-mile cross-state route is now pegged at between $7.3 and $10.4 billion, according to the recent report issued by the state’s “Blue-Ribbon Panel on Hyperloop”. It’s likely to cost much more by the time they get around to building it, if they do at all, and if it actually works.

Hyperbole?

My skepticism about hyperloops is based in part on the hucksterism that often characterizes appeals for public funding of large projects, and hyperloop hucksterism has already taken place. For example, in 2013 Elon Musk estimated that a Hyperloop system would cost about $11.5 million per mile. By 2016, the mid-point estimate for a route in the San Francisco Bay Area was over $100 million per mile. A friendlier route in Dubai is expected to cost $52 million per mile. So to be conservative, we saw 5x to 10x higher costs in a matter of three years. But now, Virgin One says it can construct a route in Missouri for less than the per-mile cost of the Dubai line. Well, the state Department of Transportation already owns the rights of way over significant stretches of the route (but not everywhere because the tube must be straighter than the highway).

The hyperbolic claims for hyperloop technology include speed, projected passenger fares, and ridership. According to Innovation Origins, the so-called feasibility study for the Missouri hyperloop did not assess the technology or even address the fact that no working hyperloop has ever been built or proven at full scale over any distance longer than a kilometer or so. The consultants who prepared the “study” merely assumed it would work. No test pod within a vacuum tube has achieved more than a fraction of the promised speed. The tubes were not long enough to achieve top speeds, they say, but that raises another issue: creating near-vacuum conditions in a sizable tube over very long distances. At the Innovation Origin link above, they estimate that the Missouri tube would occupy over 1 million cubic meters of space, which is at least 30 times larger than the most expansive man-made vacuum space now in existence.

The Ride

As for the passenger experience, 30 minutes to traverse the state of Missouri would be impressive, but what about comfort? First, expanding the tube’s circumference and the girth of the pods would have a disproportionate impact on cost, so conditions might either be more cramped than the promotional photos would have you believe, or the number of passenger seats per pod might be reduced. Second, rapid acceleration from zero to 700 mph would subject humans to fairly large G-forces over several minutes. Deceleration at the end of the trip might be even worse. Negotiating even mild curves would also require reduced speed and subsequent re-acceleration to avoid uncomfortably high radial G-forces. All that means the ride could be a bit uncomfortable. That also means the average speed between Kansas City and St. Louis would be significantly less than 700 mph, especially with a stop in Columbia. G-forces might not be much of a concern for freight traffic, unless it’s fresh produce.

Safety

Then there’s the vulnerability of the system. Willis Eschenbach goes into detail on some technical problems that make the hyperloop risky, such as the pressure on the tubes themselves. It would be about 20,000 pounds per square meter of tube surface, all subject to significant thermal expansion and contraction over the course of a day, with large pods racing through joints and rounding curves. Any fault or crack at any point in the tube surface would cause catastrophic deceleration of pods along the entire length of the tube. The integrity of the pressurized pods themselves is also a safety issue. And what about an earthquake? Or a loss of control and fiery pile-up of vehicles traveling on I-70 near the tubes. Or any number of other foolish or intentional sources of damage to the tube along its route?

Throughput

One of Eschenbach’s most interesting critiques has to do with passenger throughput. Musk’s original plan called for 28-passenger pods departing every 30 seconds: 3,300 passengers per hour. That would represent a substantial addition to total cross-state transportation capacity. At full utilization (which of course is unlikely), that would exceed current estimated totals for daily travel between St. Louis, Columbia, and Kansas City. And while that capacity might reduce pressure to expand other modes, such as adding an extra lane to I-70, it would not offer an excuse to eliminate highway, rail, or airport infrastructure, nor would it eliminate the need to maintain it.

Musks’s assumption might be too optimistic, however: for safety, the time between pod departures might have to be longer. than 30 seconds. Eschenbach asserts 80 that seconds would be more reasonable, which would slash capacity by about 60% relative to Musk’s estimate. And that doesn’t account for potential bottlenecks at stops where pods must be depressurized and repressurized. And if substantially heavier freight pods are intermingled with passenger pods, as anticipated, the required intervals between departures might have to be longer.

Economics

Few large transportation projects are self-funding. Typically, user fees fail to cover operating costs, let alone capital costs. The projected fares quoted by proponents of the Missouri hyperloop are low: “cheaper than the price of gas to drive” cross-state. Perhaps we could say about $25, based on that statement. That won’t make much of a dent in the cost of construction.

The hyperloop’s economic viability for freight traffic is questionable as well, though freight traffic seems to be a fallback position among boosters when confronted with the uncertainties of passenger travel via hyperloop. The Blue-Ribbon report says the expected cost of freight via hyperloop might range from $1.40 per mile to $2.80 on the high end, putting the mid-point well above the $1.69 per mile average cost of shipping by truck. Will speed make the hyperloop a competitive alternative for shippers? In fact, freight via hyperloop might be much worse than rail or truck in solving the “last mile” problem. That’s because the speeds that are its presumed advantage also mean fewer terminals are possible. The system would have to rely as heavily on integration with other modes of transportation as any other form of long-distance carriage, and perhaps more.

The last-mile problem eats into hyperloop’s presumed environmental advantages, which are not as clear cut as its enthusiasts would have you believe. Maintaining a vacuum in a gargantuan tube will not be a low-energy proposition, nor will powering the magnetic levitation/propulsion system, with or without a vacuum. Pressurized, climate-controlled pods will require still more power, and that’s to say nothing of the energy required to fabricate one-inch thick steel cylinders, huge magnets, and the rest of the support infrastructure. Reassurances that hyperloop will be powered exclusively by “green” technologies should be taken with a grain of salt. 

Virginia Postrel believes that regulation might be the biggest threat to the success of hyperloop, though she seems a bit optimistic about the actual economics of the technology. Safety will be a major concern for regulators. The technology will be subject to common carrier rules, and there will be other hurdles at the federal, state and local levels. And what of the health effects of prolonged exposure to those powerful magnetic forces? They may be insignificant, but the question will come up and possibly litigated.

Conclusion

A hyperloop cannot be built and operated without a significant and ongoing investment of public funds. The hoped-for public-private partnership needed to build the system would require major investors, and brave investors. Promoters say the project is not unlike efforts to build the railroads in the 19th century, which must have seemed like a daunting task at the time, and one involving huge financial risk. Fair enough, but the railroads stood to benefit in that age from a huge pent-up desire to exploit distant resources. The Missouri hyperloop is not quite comparable in that respect. It might be attractive mainly as a novelty, much like the Loop Trolley. Moreover, it didn’t take long for the railroads to become desperate rent-seekers, unable to profit from their heavily-subsidized investments without further public intervention on their behalf.

The hyperloop is a truly seductive idea. It’s the sort of thing that even small government types find irresistible, but there is little doubt that taxpayers will pay dearly. It’s not clear to me that the project will create meaningful social benefits or address compelling social risks. Therefore, let’s be cautious about making huge public commitments until this technology is farther along in development and the benefits can be estimated with greater certainty.