Elon Musk recently tweeted that he has received “verbal approval” for the construction of a Hyperloop between Washington, DC and New York City.
The media have fallen all over themselves to either fellate his techno-dick, or justly mock the notion that “verbal approval” means anything, when hundreds of municipalities will have to be consulted before a route can even be approved, let alone shovels put in the ground.
But Hyperloop take-down articles based on engineering complications, construction costs, municipal approvals, etc. have been done to death. Techno-libertarian venture capitalist idiots will gladly light billions of dollars on fire to support a novel idea, or even a hundred-year-old idea (i.e. unlicensed jitneys) if it has an app (Uber). So we’ll assume funding won’t be an issue. Politicians are eager for groundbreakings and ribbon cuttings, even more so if they don’t have to spend a dime of taxpayer money, so we’ll assume approvals will be relatively quick. The engineering is reasonably sound, and construction costs are covered by your VC friends and absurd valuation. So what will actually happen when a Hyperloop opens in your city?
The year is 2023, and the DC-NYC Hyperloop has opened to a predictable amount of pomp and circumstance. Elon Musk space-teleconferences in from his privatized Mars City One, fanned with space-palm leaves by his indentured space-servants. President Donald Trump, now nearing the end of his second term, personally cuts the ribbon on the only major infrastructure project of his administration. “With the opening of this fabulous new Hyperloop,” he says, “my incredible Trump Tower is now closer to the White House than it is to my beautiful Trump Building downtown!”
Tubes carry passengers at speeds of up to 700 mph between Washington, DC and New York City, with stops in Baltimore and Philadelphia. On the first day of operation, folks queue up for blocks for the first ride on the cheap, quick, and reliable Hyperloop.
On the second day, the queues are longer. On the third day, even more so. And the fourth, and the fifth. By the sixth day wait time for pods are equal to the train trip between NYC and Washington. On the ninth, the waiting time is equal to a similar Megabus trip.
This is called “induced demand” and we more frequently talk about it as the result of new road projects, but it will certainly apply here. It happens because the sudden availability of a twenty-dollar, thirty minute, one-seat ride from DC to NYC will cause many more people to make such a trip than would previously have done so. When the Smithsonian is quicker to get to from Penn Station than the Metropolitan Museum of Art, people’s travel preferences will change. It is caused not by a lack of capacity per se, but by an increased ability for people to travel. This problem can be solved in two ways: Increase ticket pricing (or tolls), or increase capacity by wildly disproportionate amounts so that almost any amount of people can use the service without it going overcapacity.
This is a problem, because Hyperloop as currently envisioned is an inherently low-capacity system, constrained by the inherent geometry and aerodynamics of the Hyperloop tube. Because of this, Hyperloop will have to dramatically raise fares to retain its speed advantage, putting it out of reach for regular use for regular people like you and me.
To demonstrate the Hyperloop capacity problem, I’m going to take some statistics from the Hyperloop whitepaper at face value, as I am not very familiar with the aerodynamics of an evacuated tube. On page twenty-five, it states that “It is critical to the aerodynamics of the capsule to keep [the ratio of tube area to capsule cross-sectional area] as large as possible, even though the pressure in the tube is extremely low” and suggests a ratio of thirty-six per cent for Hyperloop Alpha.
Since then, Elon Musk has set up “The Boring Company,” whose FAQ page states “Reducing the diameter in half reduces tunneling costs by 3–4 times” and suggests a tunnel diameter of fourteen feet for a vehicle sled or Hyperloop tunnel.
This results in a tunnel approximately 150 square feet in cross-sectional area, with a Hyperloop pod approximately fifty-five square feet in cross-sectional area, or eight feet three inches (or so) in diameter. The length of the pod is not mentioned in the whitepaper, but it is acknowledged that since the Hyperloop pod must displace its own volume the pod should be kept as short as possible — Hyperloop One (a separate company from Elon Musk’s current endeavor) is currently aiming for a twenty-eight-foot-long pod which seats twenty-eight passengers, the same as the stated capacity of the whitepaper.
To put this in perspective, some of the smallest mass-transit vehicles in operation today run on the Glasgow Subway, whose cars are seven and a half feet wide, eight and a half feet tall, forty-one feet long, and seat 122 passengers in transverse seating with 165 standees.
Due to significant g-forces inherent in Hyperloop’s design it is unlikely that standees will be allowed, nor will transverse seating be practical, limiting capacity. Additionally, the Glasgow Subway’s cars have no major mechanical devices which require the full cross-sectional area of the vehicle, whereas Hyperloop pods must have a very large compressor at the front to reduce air resistance by redirecting large quantities of low pressure air through the vehicle and out its rear. They also require large battery banks to power the vehicle while suspended in the tube. Thus, twenty-eight seated passengers (seven rows of four) in a vehicle seems like a reasonable capacity, though it is about a tenth of the capacity of the Glasgow Subway’s cars, and Hyperloop pods cannot be coupled together into a train.
Hyperloop intends to make up for deficiencies in vehicle capacity through very high frequency — a pod every two minutes is suggested. This results in thirty pods per hour, or 840 passengers per hour. Musk’s whitepaper also indicates that for periods of peak demand, pods may be scheduled thirty seconds apart, resulting in a short-term capacity of 3360 passengers per hour.
If we postulate that the tunnel operates at an average frequency of thirty pods per hour for twenty-four hours (since we can assume there will be more pods running at peak hours but fewer late at night) this is 20,160 passengers per day, or approximately 7.3 million passengers per year per tube (for a total of 14.6 million), ignoring holidays and weekends when service might be reduced, or periods where a tunnel is out of service for maintenance.
To be generous to Hyperloop, I will compare it at its best, quoted above, to the worst and most congested bottleneck in the existing Northeast Corridor: the 110 year-old two-track North River Tunnels between NYC Penn Station and New Jersey. This post will not cover the potential for Hyperloop to carry freight, or vehicles — this may be covered in a future post.
A current Northeast Regional train has eight coaches, six of which seat seventy-two passengers (Coach Class), one of which seats sixty passengers (Business Class), and one of which seats fifty passengers (Café Car) for a total of 542 passengers per train. Commuter rail equipment carries many more people: a typical New Jersey Transit bilevel commuter train has nine coaches, seating 1223 passengers in total.
The existing North River Tunnels, the primary capacity constraint on the Northeast Corridor, can theoretically handle twenty-three trains per hour per direction if all is well — a capacity of 28,129 seated passengers per hour per tunnel if only commuter trains enter, but in practice somewhat lower since some trains are Amtrak trains — at most four trains per hour according to current schedules, giving a capacity of 25,405 seated passengers per hour per tunnel. This capacity does not account for standees, of which there are many on peak-hour NJT trains. Trains may also be lengthened by several cars, allowing for more passengers per train. (Additionally, train per hour capacity is expected to increase significantly with the construction of two additional North River Tunnels under Amtrak’s Gateway Project.) Therefore, it is reasonable to say that the North River Tunnels in ideal operating conditions can (and do) easily carry more people per tunnel in one hour than the proposed Hyperloop could transport in one day.
Per the National Association of Railroad Passengers, the Northeast Regional carried 8.26 million passengers in FY16, while the Acela carried 3.44 million passengers in FY16, though this includes service to Boston and Newport News. New Jersey Transit’s statistics include only total rail network ridership, which was 90 million passengers in FY16, which we can assume at least a slim majority of whom pass through the North River Tunnels. Hyperloop’s 14.6 million passenger per year maximum capacity (assuming two tubes, one in each direction) is just over Amtrak’s current ridership in the Northeast Corridor, and about a sixth of New Jersey Transit’s.
Why include commuter trains in this analysis if Hyperloop is intended to be an inter-city mode of transportation? On Hyperloop, travel times between Manhattan and DC will be equal or shorter than travel times between Manhattan and Perth Amboy, NJ; or Elizabeth, NJ; or New Rochelle, NY; or Hicksville, NY; or even most of Queens and the Bronx. Thus, it is reasonable to expect a certain amount of “induced demand” for Hyperloop as a method of daily commuting, especially when travel time from West Baltimore to Manhattan is similar to that as from Williamsburg, Brooklyn.
The practical result of this will be an enormous amount of demand for very few slots on the Hyperloop. This will result in either long lines, wait-lists for tickets, or high ticket prices. Given that Hyperloop One’s current marketing slogan is “We’re not selling transportation, we’re selling time” we can assume that long lines or waitlists will not be desirable to the Hyperloop operators or passengers, and high ticket prices will be the norm. Precisely where the break-even will be, I cannot say, but it will be well above the twenty-dollar fare Elon Musk suggests in his whitepaper.
What does a novel transportation system, with a very high speed but very low capacity and very high ticket prices, mean for the rest of us? Hyperloop tickets will be out of reach for the poor and likely the middle class, including most residents in Philadelphia or Baltimore. Meanwhile, poor neighborhoods in Philadelphia and Baltimore will be well within the reach of well-to-do New Yorkers and Washingtonians. Those same neighborhoods will be orders of magnitude cheaper to live in than SoHo and Williamsburg, or Georgetown and Capitol Hill. If the cost of a ticket from Baltimore to NYC each day is less than the difference in rent between West Baltimore and Greenwich Village, and the time difference is negligible, folks are going to take advantage of it and move. Baltimore and Philadelphia will effectively become close-in suburbs of Washington and NYC, and be just as unaffordable, within a few years.
For those enjoying low costs of living in Philly and Baltimore, life will change rapidly. People who thought they were gentrifiers will be double-gentrified out of the neighborhoods they were gentrifying. Cheeseburgers will cost twenty-three dollars. Twelve ounce pours of Budweiser will be eight dollars. Local bodegas will be replaced by national chain bodegas. Your neighbor won’t know what Wawa is, and call hoagies “subs.” Everyone you meet will be a Yankees fan. Mass hysteria!
The practical result of the proposed low-capacity-high-price Northeast Hyperloop, then, will be this: Quick and convenient transportation for the rich, while the poor and middle class who have the misfortune of living twenty minutes from a new Hyperloop stop get the boot.
What, then, could an equitable Hyperloop look like? The answer: very expensive. As previously quoted by The Boring Company, increasing the tunnel diameter by a factor of two will increase construction costs by four times, and since the hyperloop pod is recommended to only occupy thirty-six per cent of the tunnel’s cross-sectional area, the pod may only be sixteen feet in diameter, about the width of a smaller wide-body jet airliner. (It will also take much, much more energy to accelerate and stop — but this is not part of today’s analysis.) This might be able to fit an additional four rows of seating, perhaps six if the designers are creative with double-decker seating, resulting in a 100-or-so seat Hyperloop pod, depending on its length. This results in a capacity of 3000 passengers per hour at two-minute headways, or 12,000 passengers per hour at thirty-second headways (assuming that is practical given that 100 people must exit and enter the pod within that headway.) This is still less than half the capacity of the North River Tunnels, so even larger tubes and vehicles will be needed to match even existing service levels — and it likely still won’t be enough, because we have absolutely no idea, nor a good way of predicting, the magnitude of demand a thirty-minute ride from NYC to DC might induce.
An equitable Hyperloop with affordable tickets for all is almost certainly not possible with private financing, in short. The public, in turn, has no business financing unproven technology. Elon Musk is correct to label Hyperloop a “fifth mode of transport” — it simply cannot replace boats, trains, cars, or planes. Though it may be economically feasible, physically-sound, and politically doable, Hyperloop fails on the one metric which drove Elon Musk to create it: it cannot and will not be able to transport enough people to reduce traffic or save travel times, at least for regular folks.
In the current political climate of late capitalism, with deferred maintenance the norm, infrastructure investments from the government few and far between, and our basic systems crumbling around us, it’s tempting to look to the private sector for a magic bullet. Hyperloop is one of those, as are self-driving cars, and the “sharing economy.” But even if the private sector can offer a real magic bullet, how do you know the magic gun isn’t aimed squarely at you? There are studied, practical regional projects out there right now — NEC Future, the Gateway Tunnels, California High Speed Rail, and hundreds of local projects to increase train frequencies and speeds, which can be implemented for a fraction of the cost and which will benefit millions more riders.
We cannot afford to wait around for the private sector to save us. It’s time to for the government (and the government alone) to invest in new trains, subways, light rail systems, and buses; and in maintaining our existing roads and bridges; no matter what kind of magic innovation the techno-libertarian dreamers say they can bring us.
