Note: this post is secretly about Hyperloop and Elon Musk’s most likely fraudulent claim about the Northeast Corridor. But it’s an interesting discussion more in general. Not all such technology is vaporware the way Musk’s efforts are. See more on The Boring Company’s false claims in a piece I published at Urbanize.LA a few days ago.
The upper limit of conventional high-speed rail seems to be 360 km/h. In Japan, experiments at that speed have succeeded, but there already are problems with noise, stopping distance, and catenary wear, and currently trains top at 320; plans to go at 360 depend on a future Shinkansen extension to Sapporo. In China, the maximum speed is 350, with trains capable of reaching 380 but not doing so in practice. In Continental Europe the maximum speed for new lines is 320-330 km/h, whereas in Britain HS2 is designed for about 350 km/h (220 mph).
Faster technologies exist, in service, today. Shanghai’s Transrapid tops at 431 km/h in service, and JR Central’s under-construction maglev line is targeted at 500 km/h, with tests at 600. Vactrains can go even faster, but are still untested technology (and this includes Hyperloop variants). The question is, where is there room for such technology? So far, Siemens’ attempts to sell Transrapid failed beyond the Shanghai airport connector, an orphan 30 km line going from the airport to the edge of the built-up area of the city. JR Central is building the Chuo Shinkansen maglev between Tokyo and Osaka, but so far there are no plans to extend this technology elsewhere – even within Japan, the state is continuing with building Tokyo-Sapporo as conventional Shinkansen rather than maglev.
The Tokyo-Osaka line is somewhat sui generis. JR Central is currently running about 14 trains per hour at the peak on the Tokaido Shinkansen between Tokyo and Shin-Osaka, each with 1,323 seats, and they’re generally full. It is also old – as the first HSR line in the world, it has a curve radius of 2.5 km (newer lines start at 4 km and go up), and a top speed of 270 km/h. This is exactly the sort of situation that favors new technology. The Tokaido Main Line was a popular intercity line in the late 1950s, but Japan National Railways couldn’t add more express trains without bumping against the capacity limit imposed by slower trains using the line; this tilted it in favor of building the Shinkansen. The Paris-Lyon main line was similarly busy in the 1970s, encouraging the construction of the LGV Sud-Est as a bypass. Nowhere in the world except Tokyo-Osaka is there a full conventional HSR line, except Paris-Lyon – but see later why it is a poor candidate for faster technology.
The main tradeoff with maglev, or even faster technology, is cost. This comes from two places. First, higher top speed requires much more advanced civil engineering, with wider curves, which means more tunnels and viaducts. Conventional HSR can limit costs by climbing steeper grades than legacy trains (the LGV Sud-Est has no tunnels, the legacy Paris-Lyon line does). Maglev can climb even faster grades, but once the speed crosses into the vactrain range, the vertical curve radius required to achieve a steep grade is so wide that it is no longer possible to vertically hug terrain the way European HSR lines do.
The second place is the urban approaches. In theory, this should be a strength of faster-than-conventional rail technology, which has a lower minimum curve radius than HSR at equal speed. But in practice, conventional HSR can leverage existing railroad lines on the urban approaches. At lower speed the stopping distances are shorter, so capacity is higher; the upper limit at speed maybe 12-15 trains per hour, but on a low-speed approach it’s closer to 24-30, so it’s possible to share tracks with legacy commuter and intercity trains.
In Japan, Spain, and Taiwan the HSR track gauge is different from the legacy gauge, so track-sharing is not possible in the major cities, driving up the cost of urban approaches. In smaller cities, Japan and Spain have gauge-change technology, which takes too much time to be of use in capacity-constrained big cities but can allow track sharing on branches. But unconventional technology cannot share tracks anywhere, requiring tunnels on urban approaches. The cost of 20 km of urban tunnel can easily match that of 200 km of at-grade greenfield HSR outside urban areas. The Chuo Shinkansen’s cost, around $200 million per km, comes from the fact that 70-80% of the line is underground, in urban areas and under mountains.
This implies that unconventional technology is most useful when there is limited benefit to be gained from track sharing. This includes the following situations:
- The cities served do not have usable legacy rail approaches, or else have a surplus of space within which to build a new approach.
- There is no need to branch and use legacy track at lower speed.
- There is no preexisting high-quality track that HSR can use, either at high speed outside cities or at medium speed on approaches.
In North America, FRA regulations traditionally led to situation #1. But FRA regulations seem to be changing, which makes track-sharing on approaches more feasible; practically every city has approaches with a surplus of passenger rail capacity (yes, even New York – Amtrak runs 4 trains per hour into Penn Station from the west at the peak, it just uses these slots poorly). In Europe, cities with poor approaches are more likely to be served on a branch, since the rest of the network is so strong. Situation #2 never applies here – branching is always useful, letting the LGV Sud-Est carry not just Paris-Lyon trains but also Paris-Marseille, Lille-Lyon, London-Lyon, Paris-Geneva, etc.
Some of the stronger intercity travel markets are in situation #3, but most aren’t. In North America, the Northeast Corridor has long stretches of high-quality track, either already capable of high speed or capable with a small number of curve modifications. That characteristic alone makes it exceptionally bad for unconventional rail technology: such technology would need a new alignment through hundreds of kilometers of suburbia in Massachusetts, Rhode Island, New Jersey, Pennsylvania, and Maryland. Toronto is also a poor candidate for unconventional technology, since it has a long stretch of suburbia in both directions with high-quality four-track commuter rail, straight enough for 200 km/h or even more. Significant suburban tracks are also useful in California (Caltrain, parts of Metrolink) and Chicago. Only the Pacific Northwest, Portland-Seattle-Vancouver, has a real shortage of usable legacy track even on the approaches. So is it a good candidate for unconventional technology? No, for reasons of distance.
The optimal distance
Faster-than-conventional rail is silly at short distance. The difference in travel time is smaller and does not justify the expense. Access and egress times are fixed, and may even go up if the station locations are less central (the Chuo Shinkansen won’t serve Tokyo Station but rather Shinagawa, a few km south of the CBD). So focusing on in-vehicle time is less useful. The Chuo Shinkansen is really at the lowest end of what is acceptable. It works because, again, the Tokaido Shinkansen is at capacity. Tokaido is also relatively circuitous in order to avoid mountains – the distance from Tokyo to Shin-Osaka is 515 km on the Tokaido Shinkansen, 438 on the Chuo Shinkansen, and 405 on a straight line. On the Northeast Corridor, the New York-Washington distance is 362 km on the railroad and 330 on a straight line, a much smaller difference.
Conversely, faster-than-conventional rail is questionable at very long distance. At maglev speed, a New York-Los Angeles train would take perhaps 12 hours, not really competitive with planes for people who don’t mind flying. At vactrain speed, the train would be competitive. However, in either case, trains require linear infrastructure, and repackaging them as a new Hyperloop doesn’t change this basic fact. Ignoring the effects of terrain, a 4,000 km vactrain or maglev line costs ten times as much as a 400 km line. This is not the case for air travel, which requires no fixed infrastructure between the airports.
There should be a good zone in the middle, say the 1,000-1,500 km range. This includes city pairs like Beijing-Shanghai, New York-Chicago, Tokyo-Sapporo, Tokyo-Fukuoka, Delhi-Mumbai, Delhi-Kolkata, and some international European pairs like Paris-Madrid. Going up to 2,000 there are also New York-Miami, Chicago-Dallas-Houston, Beijing-Guangzhou, and Los Angeles-San Francisco-Seattle; in China, where conventional HSR is faster, even 1,000-1,300 km is well within conventional HSR capabilities (Beijing-Shanghai is 1,300).
However, the fact that there is this sweet spot for unconventional rail does not mean that the construction costs are affordable. This remains a question mark. Maglev costs are either in line with HSR costs at equal tunnel proportion, or somewhat higher. The Shanghai maglev cost 10 billion RMB in 2003, which in PPP terms is maybe $100 million per km for an elevated suburban/exurban line (bad, but not terrible), and in exchange rate terms (imported technology) is somewhat more than half that. The Chuo Shinkansen seems to be $200 million per km, 70-80% underground, which is in line with urban tunneling costs in Japan but high by the standards of exurban tunneling (the 60% tunneled extension of the Tohoku Shinkansen to Shin-Aomori was $55 million per km).
The upshot is that a New York-Chicago maglev is likely to cost like 1,200 km of HSR. The western half of this line is easy – maybe a short tunnel in suburban Chicago is required, but there’s so much right-of-way space that an above-ground urban approach should be fine. The eastern half of this line consists of 600 km of pain in the Appalachians, suburban New Jersey, and a new tunnel under the Hudson. Costs approaching $100 billion are likely, and I don’t know that the benefits are commensurate.
Can you start big?
A short maglev or vactrain is of little use. Given the expense of approaches, the best use of expensive infrastructure may well be to build multiple lines using the same approach. For example, not just New York-Chicago or New York-Atlanta-Miami, but both at once, to take advantage of the same maglev tunnel under the Hudson. By itself New York-Chicago might be good enough, but it’s unclear – it’s nowhere the huge benefit/cost ratio coming from a program for conventional HSR on the Northeast Corridor at normal first-world rates.
I think this is the biggest risk with unconventional rail technology. Its basic characteristics suggest that there should be a distance range at which it works well – not too short so as to offer too little benefit versus conventional HSR, not too long so as for construction costs to grind it down. But it’s equally possible that the two bad zones, too short and too long, really overlap, so that 1,200-km lines are still too expensive to compete with planes while not offering enough speed benefit over conventional HSR to justify all this new construction.
The problem, then, is that it’s difficult to start big with a risky technology. The shortest useful maglev segment, Tokyo-Nagoya, is still well over $50 billion, and Tokyo-Osaka approaches $100 billion. This is on a route with proven demand; what about routes that don’t parallel overcrowded conventional HSR? Some government will need to take a $100 billion gamble on a long route hoping that the 1,200-km niche really exists.
I’d been making cryptic remarks about a possible job offer for a month, and a week ago I tweeted when I heard the final no. I didn’t want to say where I was interviewing until after I heard back, either way; now that I have, I’d like to talk more about the process, and what I think it means for transportation criticism in general.
A few weeks after I posted that I’m transitioning to working in transit or transit writing full-time, a recruiter reached out to me. I wouldn’t have applied myself, not out of ideological opposition to working on Hyperloop, but because until that point, I imagined they wouldn’t have wanted me working there anyway. But once the recruiter emailed me, I started the interview process. It went well. The company was familiar with my criticism of the initial concept and of startups’ own attempts to build it (the last link is Hyperloop One, the one before it is a different company). We talked about the technology, about which models I’d use to evaluate it, about various ways the system could be made more convenient.
People who are familiar with the interview process in the tech industry know that it is long and laborious. There are multiple rounds of interviews, with multiple people involved. Programming jobs involve something called whiteboarding, in which the interviewer will ask the interviewee to solve a coding problem on a whiteboard. I’m not a programmer, unless one counts QBASIC as programming, so I didn’t do any whiteboarding, but the same concept of interview meant there were a lot of hard on-the-spot technical questions. (In contrast, when I interviewed at Frontier, there were hard on-the-spot questions about political and social trends.)
Where I got stuck was American immigration policy. In the US, unlike in normal countries like Canada or Singapore or France, the skilled work visa process is based on a hard cap on the number of visas (called H-1B), rather than on a minimum salary requirement or a labor market analysis to make sure there are more jobs than qualified citizens, both of which criteria are easy to meet in tech. The H-1B cap is too tight – it’s oversubscribed by a factor of about 2; earlier this decade there was political consensus in the US elite that it needed to be lifted, but partisan politicking prevented this from happening. By mid-decade, even before Trump, the consensus frayed, thanks in no small part to anti-immigration reform conservatives, especially Reihan Salam (and, within the urbanist sphere, Aaron Renn). Academia and nonprofit research organizations, such as Frontier (or TransitCenter, or RPA), are exempt from the cap. Tech firms aren’t. This imposes a queue for getting a visa; HR at Hyperloop One said it would be a year, I think it would’ve been a year and a half. It took about a month to figure out whether Hyperloop One could work with me as a remote outside contractor, and when they realized they couldn’t, they had to tell me they couldn’t hire me.
My impressions of Hyperloop’s current status
Elon Musk’s original writeup was a scribble. Very little about it was salvageable. Hyperloop One is more serious. I believe that the most quotable criticism I made of the project in 2013 – the “barf ride” line – is being solved. As I said in 2013, I believe it is not too hard to solve the basic problem of curve radii; the problem is that it makes the civil engineering more expensive, by requiring more tunnels and more viaducts.
We didn’t discuss construction costs at the interview. I think of this as a point in the company’s favor, actually; they’d know that my understanding of construction costs is at too high a level, useful for policymakers but not for actual consultants or contractors. A few months ago, before this process started, I read somewhere that the company says Hyperloop would be 2/3 as expensive as conventional high-speed rail per km, up from Musk’s laughable 1/10 estimate. I’m skeptical about 2/3, but I’m willing to say “I’ll believe it when I see it” and not “yeah, right.”
The capacity constraints coming from the narrow tube diameter are also a problem that I think the company is capable of solving; the cost of a wider tube is higher, but in far less than linear proportion to the extra capacity provided.
There remain two big classes of hitches, one technical and one economic. The technical hitches involve materials engineering that I don’t understand as well, regarding sway inside the tube, ground subsidence, and construction tolerances. I am channeling other critics here; some of them are experts in the field and I am inclined to trust them. I’ve always taken these issues as a black box for conventional HSR and even 500-600 km/h service (maglev or conventional – the TGV reached 574 km/h in an experiment with a special train with a higher power-to-weight ratio), but at higher speeds, they become more serious.
My default assumption is that it’s still solvable at 1000+ km/h, but requires more delicate engineering, which may drive up construction costs even further. Even in my initial writeup I was implicitly arguing the required delicate engineering was such that it was inappropriate to generalize from the costs of oil pipelines, rather than from those of maglev. But it’s possible that the required materials and safety engineering will lead to much higher construction costs, and it’s possible that more basic research is required before it’s viable.
The economic hitch is, what is Hyperloop for? The technology suffers from tension between two opposing forces. The first force is speed: as a very fast technology, Hyperloop is the most useful for long-distance travel. At the distance of Musk’s original Los Angeles-San Francisco idea, security theater and design compromises about station locations (Sylmar and the East Bay, originally) would eat up the entire travel time advantage over conventional HSR. At longer distance, such as New York-Chicago, Hyperloop would still win on time, just as planes beat HSR on time on corridors in the 1,000 km range today. The second force is that Hyperloop still requires linear infrastructure, so it becomes less cost-effective versus planes as the distance increases.
Hyperloop One is a consulting firm. I was asked at the interview about the technology’s applicability in multiple geographies, and gave my opinions (“this place is a good candidate, that place isn’t”). So the company can’t just up and decide on an initial segment, which should probably be a connection from New York (probably in Jersey City or Hoboken) to either South Florida or Chicago. Complicating things, such an initial segment would require many tens of billions of dollars of capital investment, which is not easy for a startup to do. There’s a real problem with using the tech startup model to develop capital-intensive infrastructure, and it’s possible such vactrain technology will always fall between the conventional HSR and airplane chairs. I for one will keep putting vactrains in my 22nd-century science fiction, but not in my near-future science fiction.
One of the lines I wrote in my initial post is that tech megalomaniacs believe that “people who question [the entrepreneur] and laugh at his outlandish ideas will invariably fail and end up working for him.” I recognize the irony in my almost-working for Hyperloop One.
And yet, I think it offers a valuable lesson about what I variously call sycophancy, or a courtier mentality. I mentioned this about the tech press in the first post; the national political press is less sycophantic (since it can be loyal to an opposition party or political faction, and can draw on the opposition for criticism of current leadership). But local political actors in areas without real political opposition can act like royal courtiers at times, unreasonably praising the leader and begging for scraps. I’ve criticized the RPA for this, for example here: Governor Andrew Cuomo proposed a new airport connector with negative transportation value, and while the area’s transit bloggers all said no, the RPA studied the idea seriously.
The connection with Hyperloop is that I hit the concept pretty hard, and still would’ve been hired but for the US’s broken immigration policy. I don’t know if it’s generalizable to tech. I know it is true in math academia, where if I make a serious criticism of someone’s research program, it’s quite likely we will then write a paper together. For example, my advisor formulated a conjecture he called Dynamical Manin-Mumford; two professors, Rochester’s Tom Tucker and UBC’s Dragos Ghioca, later my own postdoc advisor, found a counterexample, and wrote it up together with my advisor. Nowadays the different researchers in the field are trying to prove different weaker versions of the conjecture that might still be true.
This collaborative aspect is certainly true of transit blogging. I spend a lot of time talking about transit with my biggest critic, who argues my argument about construction costs is spurious and the US is only expensive due to inexperience; I also talk a lot to people who are more nitpickers than critics, like Threestationsquare. I’ve seen the same sentiment at a thinktank whose founder I criticized years ago, and my understanding is that the RPA too is familiar with my writings. But I don’t know if it’s true of government hiring as much – if the MTA, let alone anyone working for Cuomo, is interested in hiring a critic; but then again, MTA hiring has severe problems.
Still, I’d draw a lesson and tell people who write about transportation to be less afraid of being critical. It’s a natural fear; I have it too, when I have criticism for a blogger or Twitter user who I know or consider part of my in-group. But the only result of suppressing criticism is that people who have bad ideas keep promulgating them and either never realize they’re wrong (if they’re honest) or keep acquiring suckers (if they’re dishonest). People who are interested in better transportation recognize this and seek out the critic. Megalomaniacs who are interested in selling themselves suppress and ignore the critic. We know which side Hyperloop One is on; but where is New York’s political system?
The future of my work
I can’t legally work in the US, unless it’s for a cap-exempt institution, which means either a university (that ship sailed five months ago) or a thinktank. Canada is looking unlikely – a consultancy I applied for ended up hiring someone else they felt was more qualified, and Metrolinx isn’t going to hire me. My French is conversational, but not good enough to apply for Keolis’s planning positions here, of which they have plenty, including some I’m otherwise qualified for.
This means I’m going to do transportation writing full-time for the foreseeable future. My plan is to invest in this blog more to make it look nicer (two pieces I’ve recently sent out have decent graphics), and (almost certainly) start a Patreon account in which people who pitch in a few dollars a month can influence what I write about. My intention is to commit to a post every week, not counting personal stuff like this post. I don’t expect this to net me a lot of money, but together with freelancing income, it should be enough to live on in a developed country with universal health care.
In 2009, studies began for a replacement of the Baltimore and Potomac (B&P) Tunnel. This tunnel, located immediately west of Baltimore Penn Station, has sharp curves, limiting passenger trains to about 50 km/h today. The plan was a two-track passenger rail tunnel, called the Great Circle Tunnel since it would sweep a wide circular arc; see yellow line here. It would be about 3 kilometers and cost $750 million, on the high side for a tunnel with no stations, but nothing to get too outraged about. Since then, costs have mounted. In 2014, the plan, still for two tracks, was up to $1 billion to $1.5 billion. Since then, costs have exploded, and the new Final Environmental Impact Statement puts the project at $4 billion. This is worth getting outraged about; at this cost, even at half this cost, the tunnel should not be built. However, unlike in some other cases of high construction costs that I have criticized, here the problem is not high unit costs, but pure scope creep. The new scope should be deleted in order to reduce costs; as I will explain, the required capacity is well within the capability of two tracks.
First, some background, summarized from the original report from 2009, which I can no longer find: Baltimore was a bottleneck of US rail transportation in the mid-19th century. In the Civil War, there was no route through the city; Union troops had to lug supplies across the city, fighting off mobs of Confederate sympathizers. This in turn is because Baltimore’s terrain is quite hilly, with no coastal plain to speak of: the only flat land on which a railroad could be easily built was already developed and urbanized by the time the railroad was invented. It took until the 1870s to build routes across the city, by which time the US already had a transcontinental railroad. Moreover, intense competition between the Pennsylvania Railroad (PRR) and the Baltimore and Ohio (B&O) ensured that each company would built its own tunnel. The two-track B&P is the PRR tunnel; there’s also a single-track freight tunnel, originally built by the B&O, now owned by CSX, into which the B&O later merged.
Because of the duplication of routes and the difficult geography, the tunnels were not built to high standards. The ruling grade on the B&P is higher than freight railroads would like, 1.34% uphill departing the station, the steepest on the Northeast Corridor (NEC) south of Philadelphia. This grade also reduces initial acceleration for passenger trains. The tunnel also has multiple sharp curves, with the curve at the western portal limiting trains today to 30 mph (about 50 km/h). The CSX tunnel, called Howard Street Tunnel, has a grade as well. The B&P maintenance costs are high due to poor construction, but a shutdown for repairs is not possible as it is a key NEC link with no possible reroute.
In 2009, the FRA’s plan was to bypass the B&P Tunnel with a two-track passenger rail tunnel, the Great Circle Tunnel. The tunnel would be a little longer than the B&P, but permit much higher speeds, around 160 km/h, saving Acela trains around 1.5 minutes. Actually the impact would be even higher, since near-terminal speed limits are a worse constraint for trains with higher initial acceleration; for high-performance trains, the saving is about 2-2.5 minutes. No accommodation was made for freight in the original plan: CSX indicated lack of interest in a joint passenger and freight rail tunnel. Besides, the NEC’s loading gauge is incompatible with double-stacked freight; accommodating such trains would require many small infrastructure upgrades, raising bridges, in addition to building a new tunnel.
In contrast, the new plan accommodates freight. Thus, the plan is for four tracks, all built to support double-stacked freight. This is despite the fact that there is no service plan that requires such capacity. Nor can the rest of the NEC support double-stacked freight easily. Of note, Amtrak only plans on using this tunnel under scenarios of what it considers low or intermediate investment into high-speed rail. Under the high-investment scenario, the so-called Alternative 3 of NEC Future, the plan is to build a two-track tunnel under Downtown Baltimore, dedicated to high-speed trains. Thus, the ultimate plan is really for six tracks.
Moreover, as pointed out by Elizabeth Alexis of CARRD, a Californian advocacy group that has criticized California’s own high-speed rail cost overruns, the new tunnel is planned to accommodate diesel trains. This is because since 2009, the commuter rail line connecting Baltimore and Washington on the NEC, called the MARC Penn Line, has deelectrified. The route is entirely electrified, and MARC used to run electric trains on it. However, in the last few years MARC deelectrified. There are conflicting rumors as to why: MARC used the pool of Amtrak electric locomotives, and Amtrak is stopping maintaining them as it is getting new locomotives; Amtrak is overcharging MARC on electricity; MARC wants fleet compatibility with its two other lines, which are unelectrified (although the Penn Line has more ridership than both other lines combined). No matter what, MARC should immediately reverse course and buy new electric trains to use on the Penn Line.
Freight trains are more complicated – all US freight trains are dieselized, even under catenary, because of a combination of unelectrified yards and Amtrak’s overcharging on electric rates. However, if freight through the Great Circle Tunnel is desired, Amtrak should work with Norfolk Southern on setting up an electric district, or else Norfolk Southern should negotiate trackage rights on CSX’s existing tunnel. If more freight capacity is desired, private companies NS and CSX can spend their own money on freight tunnels.
In contrast, a realistic scenario would ignore freight entirely, and put intercity and regional trains in the same two-track tunnel. The maximum capacity of a two-track high-speed rail line is 12 trains per hour. Near Baltimore Penn the line would not be high-speed, so capacity is defined by the limit of a normal line, which is about 24 tph. If there is a service plan under which the MARC Penn Line could get more than 12 tph at the peak, I have not seen it. The plans I have seen call for 4 peak tph and 2 off-peak tph. There is a throwaway line about “transit-like” service on page 17, but it’s not clear what is meant in terms of frequency.
Regardless of what the state of Maryland thinks MARC could support, 12 peak regional tph through Baltimore is not a reasonable assumption in any scenario in which cars remain legal. The tunnels are not planned to have any stations, so the only city station west of Baltimore Penn is West Baltimore. Baltimore is not a very dense city, nor is West Baltimore, most famous for being the location of The Wire, a hot location for transit-oriented development. Most of Baltimore’s suburbs on the Penn Line are very low-density. In any scenario in which high-speed rail actually fills 12 tph, many would be long-range commuters, which means people who live in Baltimore and work in Washington would be commuting on high-speed trains and not on regional trains. About the upper limit of what I can see for the Penn Line in a realistic scenario is 6 tph peak, 3-4 tph off-peak.
Moreover, there is no real need to separate high-speed and regional trains for reasons of speed. High-speed trains take time to accelerate from a stop at Baltimore: by the portal, even high-acceleration sets could not go much faster than 200 km/h. An in-tunnel speed limit in the 160-180 km/h area only slows down high-speed trains by a few seconds. Nor does it lead to any noticeable speed difference with electrified regional trains, which would reduce capacity: modern regional trains like the FLIRT accelerate to 160 km/h as fast as the fastest-accelerating high-speed train, the N700-I, both having an acceleration penalty of about 25 seconds.
The upshot is that there is no need for any of the new scope added since 2009. There is no need for four tracks; two will suffice. There is no need to design for double-stacked freight; the rest of the line only accommodates single-stacked freight, and the NEC has little freight traffic anyway. Under no circumstances should diesel passenger trains be allowed under the catenary, not when the Penn Line is entirely electrified.
The new tunnel has no reason to cost $4 billion. Slashing the number of tunnels from four to two should halve the cost, and reducing the tunnels’ size and ventilation needs should substantially reduce cost as well. With the potential time gained by intercity and regional trains and the reduced maintenance cost, the original budget of $750 million is acceptable, and even slightly higher costs can be justified. However, again because the existing two-track capacity can accommodate any passenger rail volume that can be reasonably expected, the new tunnel is not a must-have. $4 billion is too high a cost, and good transit activists should reject the current plan.
As the ongoing attempt to build a Hyperloop tube in California is crashing due to entirely foreseen technical problems, the company trying to raise capital for the project, Hyperloop One, is looking at other possibilities in order to save face. A few come from other passenger routes: Stockholm-Helsinki is one option, and another is the Dubai-Abu Dhabi, which looks like it may happen thanks to the regime’s indifference to financial prudence. Those plans aren’t any better or worse than the original idea to build it in California. But as part of their refusal to admit failure, the planners are trying to branch into express freight service. Hyperloop freight is especially egregious, in a way that’s interesting not only as a way of pointing out that tech entrepreneurs don’t always know what they’re doing, but also because of its implications for freight service on conventional high-speed rail.
First, let’s go back to my most quoted line on Hyperloop. In 2013 I called it a barf ride, because the plan would subject passengers to high acceleration forces, about 5 m/s^2 (conventional rail tops at 1.5 m/s^2, and a plane takes off at 3-4 m/s^2). This is actually worse for freight than for passengers, which is why the speed limits on curves are lower for freight trains than for passenger trains: as always, see Martin Lindahl’s thesis for relevant European standards. Freight does not barf, but it does shift, potentially dangerously; air freight is packed tightly in small pellets. Existing freight trains are also almost invariably heavier than passenger trains, and the heavier axle loads make high cant deficiency more difficult, as the added weight pounds the outer rail.
Another potential problem is cant. Normally, canting the tracks provides free sideways acceleration: provided the cant can be maintained, no component of the train or tracks feels the extra force. Cant deficiency, in contrast, is always felt by the tracks and the frame of the train; tilting reduces the force felt in the interior of the train, but not on the frame or in the track. At Hyperloop’s proposed speed and curve radius, getting to 5 m/s^2 force felt in the interior of the train, toward the floor, requires extensive canting. Unfortunately, this means the weight vector would point sideways rather than down, which the lightweight elevated tube structure would transmit to concrete pylons, which have high compressible strength but low tensile strength. This restricts any such system to carrying only very lightweight cargo, of mass comparable to that of passengers. This is less relevant to conventional high-speed rail and even maglev, which use more massive elevated structures, but conversely the problem of high forces on the outer rail ensures cant deficiency must be kept low.
Taken together, this means that high-speed freight can’t be of the same type as regular freight. Hyperloop One, to its credit, understands this. The managers are furiously trying to find freight – any kind of freight – that can economically fit. This has to involve materials with a high ratio of value to mass, for example perishable food, jewelry, and mail. SNCF ran dedicated TGV mail trains for 31 years, but decided to discontinue the service last year, in the context of declining mail volumes.
High-speed freight has a last mile problem. Whereas high-speed passenger service benefits from concentration of intercity destinations near the center of the city or a handful of tourist attractions, high-speed freight service has to reach the entire region to be viable. Freight trains today are designed with trucks for last-mile distribution; starting in the 1910s, industry dispersed away from waterfronts and railyards. The combination of trucks and electrification led to a form of factory building that is land-intensive and usually not found in expensive areas. Retail is more centralized than industry, but urban supermarkets remain local, and suburban ones are either local or auto-oriented hypermarkets. Even urban shopping malls as in Singapore are designed around truck delivery. The result is that high-speed freight must always contend with substantial egress time.
Let us now look at access time. How are goods supposed to get from where they’re made to the train station? With passengers, there are cars and connecting transit at the home end. There’s typically less centralization than at the destination end, but in a small origin city like the secondary French and Japanese cities, travel time is not excessive. In a larger city like Osaka it takes longer to get to the train station, but car ownership is lower because of better public transit, which increases intercity rail’s mode share. On freight, the situation is far worse: industry is quite dispersed and unlikely to be anywhere near the tracks, while the train station is typically in a congested location. Conventional rail can build a dedicated freight terminal in a farther out location (for example, auto trains in Paris do not use Gare de Lyon but Bercy); an enclosed system like Hyperloop can’t.
And if industry is difficult to centralize, think of farmed goods. Agriculture is the least centralized of all economic activities; this is on top of the fact that of all kinds of retail, supermarkets are the most local. Extensive truck operations would be needed, just as they are today. And yet, outside analysts are considering perishables as an example of a good where Hyperloop could compete.
With that in mind, any speed benefits coming from high-speed freight services vanish. There are diminishing returns to speed. Since the cost of extra speed does not diminish, there’s always a point where reducing travel time stops being useful, since the effect on door-to-door travel time is too small to justify the extra expense. The higher the total access plus egress time is, the sooner this point is reached, and in freight, the total access and plus egress time is just too long.
In passenger service, the problem of Hyperloop is that it tries to go just a little bit too far beyond conventional high-speed rail. The technical problems are resolvable, at extra cost, and in a few decades, vactrains (probably based on maglev propulsion rather than Elon Musk’s air bearings) may become viable for long-distance passenger rail.
In freight, the situation is very different. Successful freight rail companies, for example the Class I railroads in North America, China Railways, and Russian Railways, make money off of hauling freight over very long distances at low cost. Quite often this is because the freight in question is so heavy that even without substantial fuel taxes, trucks cannot compete on fuel or on labor costs; this is why Western Europe’s highest freight rail mode share is found in Sweden, with its heavy iron ore trains, and in Switzerland, Finland, Austria, with their long-distance freight across the Alps or toward Russia. Increasing speed is not what the industry wants or needs: past US experiments with fast freight did not succeed financially. The fastest, highest-cost mode of freight today, the airplane, has very low mode share, in contrast with the popularity of planes and high-speed trains in passenger service.
None of this requires deep analysis; in response to Hyperloop One’s interest in freight, an expert in logistics asked “why do we need to move cargo at 500 mph?“. The problem is one of face. The entrepreneurs in charge of Hyperloop One cannot admit that they made a mistake, to themselves, to their investors, or to the public. They are bringing the future to the unwashed masses, or so they think, and this requires them to ignore any problem until after it’s been solved, and certainly not to admit failure. Failure is for ordinary people, not for would-be masters of the universe. The announcement of the grand project is always more bombastic and always reaches more people than the news of its demise. It’s on those of us who support good transit and good rail service to make sure the next half-baked idea gets all the skepticism and criticism it deserves.
A year ago, based on a leak from Senator Charles Schumer’s office, I attacked Amtrak for paying double for its new high-speed trains – $2.5 billion for 28 trainsets, about $11 million per car. Amtrak at the time denied the press release, saying it was still in the process of selecting a bidder. However, last week Amtrak announced the new order, confirming Schumer’s leak. The trainsets are to cost $2 billion, or $9 million per car, with an additional $500 million spent on other infrastructure. The vendor is Alstom, which is branding all of its export products under the umbrella name Avelia; this train is the Avelia Liberty.
You can see a short promotional video for the trains here and read Alstom’s press release here. Together, they make it obvious why the cost is so high – about twice as high per car as that of Eurostar’s Velaro order, and three times as high as that of the shorter-lived N700 Shinkansen. The Avelia Liberty is a bespoke train, combining features that have not been seen before. Technical specs can also be seen in Alstom’s press kit. The Avelia Liberty will,
- Have a top speed of 300 km/h.
- Have articulated bogies.
- Be capable of 7 degrees of tilt, using the same system as in Alstom’s Pendolino trainset.
In particular, the combination of high speed and high degree of tilt, while technically feasible, does not exist in any production train today. It existed in prototype form, as a tilting TGV, but never made it to mass production. The Pendolino has a top speed of 250 km/h, and the ICE-T has a top speed of 240 km/h. Faster tilting trains do not tilt as much: Talgo claims the Talgo 350 is capable of lateral acceleration of 1.2 m/s^2 in the plane of the train, which corresponds to 180 mm of cant deficiency, achievable with 2-3 degrees of tilt; the tilting Shinkansen have moderate tilting as well, which the JRs call active suspension: the N700 tilts 1 degree, and appears capable of 137 mm of cant deficiency (270 km/h on 2.5 km curves with 200 mm cant), whereas the E5 and E6 tilt 2 degrees, and appear capable of 175 mm (in tests they were supposed to do 360 km/h on 4 km curves with 200 mm cant, but only run at 320 km/h for reasons unrelated to track geometry).
I have argued before, primarily in comments, that a train capable of both high speed and high degree of tilt would be useful on the Northeast Corridor, but not at any price. Moreover, the train is not even planned to run at its advertised top speed, but stay limited to 257 km/h (160 mph), which will only be achievable on short segments in Massachusetts, Rhode Island, and New Jersey. Amtrak has no funded plan to raise the top speed further: the plans for constant-tension catenary in New Jersey are the only funded item increasing top speed. There is no near-term plan on the horizon to obtain such funding – on the contrary, Amtrak’s main priority right now is the Gateway tunnel, providing extra capacity and perhaps avoiding a station throat slowdown, but not raising top speed.
Running trains at 300 km/h on the segments that allow the highest speeds today, or are planned to after the speedup in New Jersey, would save very little time (75 seconds in New Jersey, minus acceleration and deceleration penalties). Making full use of high top speed requires sustaining it over long distances, which means fixing curves in New Jersey that are not on the agenda, installing constant-tension catenary on the entire New York-Washington segment and not just over 40 km of track in New Jersey to eliminate the present-day 215 km/h limit, and building a bypass of the entire segment in southeastern Connecticut along I-95. None of these is on the immediate agenda, and only constant-tension catenary is on the medium-term agenda. Hoping for future funding to materialize is not a valid strategy: the trains would be well past the midpoint of their service lives, and spend many years depreciating before their top speed could be used.
What’s more, if substantial bypasses are built, the value of tilting decreases. In advance of the opening of the Gotthard Base Tunnel, Swiss Federal Railways (SBB) ordered 29 trainsets, without tilting, replacing the tilting Pendolino trains that go through the older tunnel. SBB said tilting would only offer minimal time reduction. The eventual cost of this order: about $36 million per trainset as long as 8 US cars. On the entire Northeast Corridor, the place where tilting does the most to reduce travel time is in Connecticut, and if the eastern half of the tracks in the state are bypassed on I-95, tilting loses value. West of New Haven, tilting is not permitted at all, because of Metro-North’s rules for trains using its tracks; on that segment, tilting will always be valuable, because of the difficulty of finding good rights-of-way for bypasses not involving long tunnels, but to my knowledge Amtrak has not made any move to lift the restriction on tilting. Even with the restriction lifted, a 300+ km/h train with moderate tilting, like the N700 or E5/6 or the Talgo AVRIL, could achieve very fast trip times, with only a few minutes of difference from a hypothetical train with the same top speed and power-to-weight ratio and 7 degrees of tilt. It may still be worth it to develop a train with both high speed and a high degree of tilt, but again, not at any cost, and certainly not as the first trainset to use the line.
Another issue is reliability. The Pendolino tilt system is high-maintenance and unreliable, and this especially affects the heavier Acela. SBB’s rejection of tilting trains was probably in part due to the reliability issues of previous Pendolino service across the Alps, leading to long delays. Poor reliability requires more schedule padding to compensate, and this reduces the advantage gained from faster speed on curves. While tilting trains are overall a net positive on curvy routes like the Connecticut segment of the Northeast Corridor, they are probably not useful in any situation in which 300 km/h top speeds are achievable for a meaningful length of time. This goes double for the Avelia Liberty, which is not a proven Pendolino but a new trainset, sold in a captive market that cannot easily replace it if there are maintenance issues.
In my post a year ago, I complained that Amtrak’s specs were conservative, and did not justify the high cost. I stand behind that assessment: the required trip times are only moderate improvements over the current schedule. At least between New York and Boston, the improvement (9 minutes plus stop penalty at New London) is less than the extent of end-of-line schedule padding, which is at least 10 minutes from Providence to Boston for northbound trains. However, to achieve these small trip time improvements, Amtrak elected to demand exacting specs from the trainsets, leading to high equipment costs.
In 2013, I expounded on this very decision by borrowing a Swiss term: the triangle of rolling stock, infrastructure, and timetable. Planning for all three should be integrated. For example, plans for increases in capacity through infrastructure improvements should be integrated with plans for running more trains, with publicly circulated sample schedules. In this case, the integration involves rolling stock and infrastructure: at low infrastructure investment, as is the case today, there is no need for 300 km/h trainsets, whereas at high investment, high top speed is required but 7-degree tilt is of limited benefit. Instead of planning appropriately based on its expectations of near-term funding, Amtrak chose to waste about a billion dollars paying double for trainsets to replace the Acela.
Amtrak’s plan for high-speed rail on the Northeast Corridor, at a cost of about $290 billion depending on the exact alternative chosen, is unacceptably costly. I went into some details of where excess cost comes from in an older post. In this post, I hope to start a series in which I focus on a specific part of the Northeast Corridor and propose a cheaper alternative than what the NEC Future plan assumes is necessary. The title is taken from a post of mine from four years ago; since then, the projected costs have doubled, hence the title is changed from 90% cheaper to 95% cheaper. In this post, I am going to focus on untangling Frankford Junction.
Frankford Junction is one of the slowest parts of the Northeast Corridor today south of New York. It has a sharp S-curve, imposing a speed limit of 50 mph, or 80 km/h. While worse slowdowns exist, they are all very close to station throats. For example, Zoo Junction just north of Philadelphia 30th Street Station has a curve with radius about 400 meters and an interlocking, so that superelevation is low. The speed limit is low (30 mph, or 50 km/h), but it’s only about 2 km out of the station; it costs about 2 minutes, and with proper superelevation and tilting the speed limit could be doubled, reducing the time cost to 25 seconds. In contrast, Frankford Junction is about 13 km out of 30th Street Station; an 80 km/h restriction there, in the middle of what could be a 200 km/h zone, makes it uneconomic for trains to accelerate to high speed before they clear the junction. This impacts about 4 km, making it a 108-second slowdown, which can be mitigated by either more tilting or a wider curve. In reality, a mixture is required.
The NEC Future plan for high-speed rail, the $290 billion Alternative 3, avoids the Frankford Junction S-curve entirely by tunneling under Center City and building a new HSR station near Market East, a more central location than 30th Street; see PDF-pp. 19, 20, and 78 of Appendix A of the environmental impact statement. This option should be instantly disposed of: 30th Street is close enough to the Philadelphia CBD, and well-connected enough to the region by public transit, that it is no worse a station choice than Shin-Osaka. The Tokaido Shinkansen could not serve Osaka Station as a through-station without tunneling; since Japan National Railways wanted to be able to extend HSR onward, as it eventually did with the Sanyo Shinkansen, it chose to serve Osaka via a new station, Shin-Osaka, 3 km away from the main station. Given the expense of long tunnels under Philadelphia, the slightly less optimal station today should be retained as good enough.
A lower-powered plan providing some HSR functionality, Alternative 2, does not include a new tunnel under Philadelphia, but instead bypasses Frankford Junction. On Appendix A, this is on PDF-pp. 19, 20, and 70. Unfortunately, the bypass is in a tunnel, which appears to be about 4 kilometers. The tunnel has to cross under a minor stream, Frankford Creek, adding to the cost. Instead, I am going to propose an alignment that bypasses the tunnel, with moderate takings, entirely above ground.
In brief, to minimize trip times without excessive construction, it is best to use the highest superelevation and cant deficiency that HSR technology supports today. The maximum superelevation is 200 mm, on the Tokaido Shinkansen (link, PDF-p. 41); there were plans to raise superelevation to 200 mm on the Tohoku Shinkansen, to permit a maximum speed of 360 km/h, but they were shelved as that speed created problems unrelated to superelevation, including noise, pantograph wear, and long braking distances. The maximum cant deficiency on existing trainsets capable of more than 300 km/h is about 180 mm, including the E5/E6 Shinkansen and the Talgo 350 and Talgo AVRIL. Tilting trains capable of nearly 300 mm cant deficiency exist, but are limited to 250 km/h so far. With 200 mm superelevation and 175 mm cant deficiency, speed in meters per second equals square root of (2.5 * curve radius in meters); the minimum curve radius for 200 km/h is then 1,235 meters.
An S-curve requires some distance to reverse the curve, to avoid shocking the train and the passengers with a large jerk, in which they suddenly change from being flung to the right to being flung to the left. If you have ridden a subway, sitting while the train was decelerating, you must have noticed that as the train decelerated, you felt some force pushing you forward, but once the train came to a complete stop, you’d be pulled backward. This is the jerk: your muscles adjusted to being pushed forward and resisting by pulling backward, and once the train stopped, they’d pull you back while adjusting back to the lack of motion. This is why S-curves built a long time ago, before this was well-understood, impose low speed limits.
With today’s computer-assisted design and engineering, it’s possible to design perfect S-curves with constant, low jerk. The limits are described in the above link on PDF-pp. 30 and 38. With the above-described specs, both sets of standards described in the link require 160 meters of ramp. For a single transition from tangent track to a fully superelevated curve, this can be modeled very accurately as 80 meters of straight track plus the circular curve (half the transition spiral is within the curve); the displacement from an actual spiral curve is small. For an S-curve, this requires double the usual transition, so 160 meters of tangent track between the two circles; bear in mind that this distance grows linearly with speed, so on full-speed 360 km/h track, nearly 300 meters are required.
Here is a drawing of two circles and a tangent track between them. The curve of course consists only of a short arc of each circle. The straight segment is a little less than 700 meters, which permits a gentle spiral. The curves have radius 1,250 meters. Takings include a charter school, a wholesale retailer, an auto shop, and what appears to be industrial parking lots, but as far as I can tell no residences (and if I’m wrong, then very few residences, all very close to industrial sites). The charter school, First Philadelphia Preparatory, is expanding, from 900 students in 2012-3 to an expected 1,800 in 2018-9. School construction costs in Pennsylvania are high, and $100 million is expected for a school of that size; see also table 5 on PDF-p. 7 here for national figures. The remaining takings are likely to cost a fraction of this one. Even with the high cost of takings, it is better to realign about 2 kilometers of track above-ground, at perhaps $150 million, than to build 4 km of tunnel, at $1.5 billion; both figures are based on cost items within the NEC Future documents. This represents a saving of about 83% over Alternative 2, which is projected to cost $116-121 billion excluding rolling stock (PDF-p. 42 of chapter 9 of the EIS).
Given the long spiral length, it may be feasible to avoid the charter school entirely. This would probably require shrinking curve radius slightly, permitting 180 or 190 km/h rather than 200 km/h. However, the travel time cost is measured in seconds: with about 11 km from the end of Zoo Junction to the northern end of Frankford Junction, of which 1 is required just to accelerate to speed, the difference between 200 and 180 km/h is 20 seconds. Further savings, reducing this time difference, are possible if the speed limit without taking the school is 190, or if trains accelerate to 200, decelerate to curve speed, and accelerate again to the north. This option would improve the cost saving over Alternative 2 to about 90%.
The correct way forward for affordable improvement of the Northeast Corridor is to look for ways in which expensive infrastructure can be avoided. If a tunnel can be replaced by a viaduct at the cost of a few extra takings, it should be. If an expensive undertaking can be avoided at the cost of perhaps 10 seconds of extra travel time, then it probably should be avoided. There should be some idea of how much it’s acceptable to spend per minute of marginal travel time saving, by segment: the New York-Philadelphia segment has the heaviest traffic and thus should have the highest maximum cost per unit of time saved. But even then, $100 million for 20 seconds is probably too high, and $100 million for 10 seconds is certainly too high.
Two years ago, when Elon Musk first proposed Hyperloop as a faster, cheaper, and more entrepreneurial alternative to California High-Speed Rail, I explained in depth what was wrong with the proposal. The curve radii were too tight for passenger comfort, and any attempt to fix them would require more expensive civil infrastructure. In general, the cost estimates in the plan were laughably low. Musk has moved on, but another team has been trying to build the system. It is planning to build a test track in the next three years, a distance of 8 km, for $150 million.
Let us analyze these costs. The per-km cost of this scheme is about $19 million, which if costs don’t run over is reasonable for HSR flat terrain, if anything a bit low. California HSR’s Central Valley segments, in more urbanized areas, are about $24-27 million/km, ex-electrification and systems (which don’t add much). This, in principle, suggests the system could be built for about the same cost as conventional HSR. Of course, it’s already far more expensive than Musk’s original estimate of $6 billion for about 650 km (including tunnels), but it still sounds like a good deal – in theory.
In practice, I’d like to go back to my often-quoted sentence in my post from two years ago, that Hyperloop would be a barf ride. The plan is to run capsules at their full speed, but only when empty. Tests with passengers would be restricted to 160 mph, or about 260 km/h. If the picture in the article describing the test track is accurate, the turn looks like its radius is perhaps 800 meters. Passengers can’t ride through this at very high speed. Even at 260 km/h, it requires full canting, and will make passengers feel noticeable extra gravitational push, about 0.2 g.
The importance of this is that any attempt to build tracks at higher speed will run into problems with both horizontal and vertical curves very quickly. The picture depicts sleek viaducts in empty land; imagine much taller viaducts, to allow the track to curve more gently than the terrain. Once the terrain becomes problematic, as it does on the approaches to the mountain crossings from the Central Valley to both the Los Angeles Basin and the San Francisco Bay Area, costs go up. This is true for any mode of transportation, up to and including mountain roads with hairpin turns, but the higher the speed, the larger the cost differential. In this situation, 4 km horizontal curve radii and 20 km vertical curve radii (about absolute minimum for conventional HSR) are expensive; 20 km horizontal curves and 230 km vertical curves are far more so. And within the urban areas, the inability of the system to leverage legacy rail tracks forces expensive urban viaducts.
Update 2016/8/16: the deal is on, per sources at Amtrak; the cost is $2.5 billion, as reported originally.
Update 9/24: as Alex Block notes in comments, sources at Amtrak deny the story, saying that Schumer spoke too soon, and there are still two bidders and Amtrak has not yet made its choice. If the cost turns out to be $1-1.25 billion rather than $2.5 billion, I will withdraw any and all criticism of the procurement process.
A press release from Senator Charles Schumer’s office is abuzz: Amtrak chose Alstom’s bid for its next order of high-speed trainsets, the Next-Generation Acelas. The press release mentions the size of the contract, $2.5 billion, and the number of jobs it would create, 750; it did not include any information relevant to passengers, such as the number of trains, the expected schedule of delivery, the expected frequency, and the expected travel time. Various media outlets have reprinted Schumer’s press release without such additional information, or indeed any analysis. Let me rectify this and provide some background as to why this order is a fleece.
The order is for 28 trainsets with 425 seats each. This can be seen here and here. Of those 28 sets, 25 should be available for maximum service, well below the 98% peak availability achieved by the TGV, but an improvement over the Acela’s current 16 trains available out of 20. There is no mention of the number of cars, which is how orders are usually priced. However, on page 30 of the technical specs, it is mentioned that the maximum length is 200 meters, equivalent to 8 cars. The capacity is equivalent to about six cars’ worth of seating at the normal seat density of economy-class HSR (including the Amtrak Regional coach), or about seven cars’ worth averaged over all occupied Acela cars. The RFP mentions half a bistro car with an option for a full car (page 21 of instructions to offerors), so eight cars per train is a reasonable assumption. I have seen references to ten cars per set, which I believe come from the option for two additional cars per train (the instructions phrase this as “an extra 33.33% capacity”). From Schumer’s press release it’s difficult to know whether the $2.5 billion figure is the base order or also the option.
Eight cars per train times 28 trains equals 224 cars. $2.5 billion divided by 224 equals $11.2 million per car; if I am wrong and these are ten-car trains, then it is $8.9 million per car. In China, a very high-speed train, capable of 350-380 km/h, costs $4 million per car; this is $900 million at the size of Amtrak’s order. In Europe, the new Eurostar order cost a total of €600-700 million for ten 16-car Velaro trainsets, about $4.7-5.5 million per car in PPP terms (see here and here); the uncertainty comes from euro:pound conversion rates and from the fact that a portion of the order is for refurbishment of the older trainsets. Siemens also sold 8-car Velaros to Deutsche Bahn for $5.2 million per car, again in PPP terms. Japanese trains are even cheaper, about $3 million per car in a recent N700 order, but only last 20 years, whereas European HSR trainsets last 40 and Amtrak specified a 30-year shelf life. The only non-US trainset order that I’ve seen that approaches the $10 million per car mark is the Velaro RUS, which is €600 million for eight 10-car trains, and this includes substantial modifications, such as winterization.
There is no excuse for such high costs. The technical specs are not particularly innovative: on page 22 of the document linked above, it is mentioned that cant deficiency should be 127 mm if the trains don’t tilt and 229 if they do, both of which figures are unimpressive by the respective standards of non-tilting and tilting trains. There is no explicit requirement for tilt. There is a requirement that trains be capable of traveling between New York and Washington in 2:21 (current trip time is 2:48) and between New Haven and Boston in 1:51 (current trip time is about 2 hours, skipping New London, which the specs require trains to stop at); there is no mention of which track upgrades are forthcoming, but given Amtrak’s heavy schedule padding, it is not difficult for a good train to meet the requirements. I do not bring these specs up to attack Amtrak for not demanding more of the trains, but to note that what Amtrak is asking is standard, so there is no reason for trains to be unusually expensive.
I will note that due to Buy America provisions, the trains will be manufactured in the US, at Alstom’s factory in Hornell. This has not caused cost blowouts for the large orders made by the New York subway, the LIRR, and Metro-North, but perhaps this order is small enough that requiring Alstom to build it at a new factory leads to major cost increases. It is also possible that due to difficulties in the bidding process, there are fewer bidders than is normal – Bombardier dropped out of the process last year, and in general, some US contracts have just one bid, with correspondingly elevated prices. But regardless of the reason, Amtrak’s order comes at a factor-of-two cost premium, and Schumer just expressed pride at the few hundred jobs that this waste would create.
The Northeast Corridor high-speed rail investment studies are moving forward, and four days ago the FRA released an early environmental impact study on the subject, as part of the NEC Future program. The study moves in part in the right direction, in that it considers many different segment-level improvements (for example, specific bypasses of curvy segments), but it still isn’t quite going in the right direction. It’s not a bad study in itself, but it does have a lot of drawbacks, and I would like to discuss the ultimate problems with its approach.
The EIS studies three alternatives, as well as an obligatory No Build option.
Alternative 1 includes minimal investment: capacity improvements already under consideration, including new Hudson tunnels; grade-separation of at-grade rail junctions, including Shell interlocking between the Metro-North New Haven Line to Grand Central and the NEC, which imposes a severe speed limit (30 mph, the worst outside major city stations) and a capacity constraint; and a limited I-95 bypass of the legacy NEC route in eastern Connecticut, to avoid the existing movable bridges. The bulk of the expense under this alternative, excluding the predominantly commuter-oriented new Hudson tunnels, involves replacing or bypassing obsolete or slow bridges with faster segments. I have advocated such an approach in certain cases for years, such as the Cos Cob Bridge; if anything, Alternative 1 does not do this enough, but I do appreciate that it uses this solution.
Alternative 2 constructs HSR along the NEC route, except for a major deviation to serve Hartford. It is also bundled with various bypasses and new stations elsewhere: under this alternative, Philadelphia and Baltimore get new stations, with extensive urban tunneling to reach those stations. Alternative 3 does the same, but considers more deviations, including a tunnel between Long Island and New Haven, and an inland route through Connecticut, closer to I-84 than to I-95 and the legacy NEC; it also constructs dedicated HSR tracks between New York and Washington.
The EIS does not include cost figures. It includes travel time figures on PDF-p. 51, which seem to be based on unfavorable assumptions: Alternative 2, called Run 5, does New York-Boston in 2:17 for trains making a few major-city intermediate stops; the Alternative 3 proposals vary widely depending on alignment, of which the fastest, the I-84 inland route, takes 1:51, again making intermediate stops.
First, the EIS includes service plan elements, stating the projected frequency of regional and express trains using the tracks. It also talks about clockface scheduling and proposes a pulse in Philadelphia, allowing timed transfers in all directions between local and express intercity trains as well as trains on the Keystone corridor. It goes further and discusses regional rail on the intercity tracks in the alternatives that include extensive new construction. In these ways, it focuses on regionwide rail integration far more than previous plans.
Second, in general, the correct way to think about NEC investment is component by component. The EIS gets closer to this ideal, by considering many different route combinations north of New York, and advancing several of them under the Alternative 3 umbrella.
And third, the concept of Alternative 1 is solid. In many cases, it is possible to bundle a trip time or capacity improvement into the replacement of an obsolete structure at very low additional cost. The example I keep coming back to is the Cos Cob Bridge, but it is equally true of the movable bridges east of New Haven. I also greatly appreciate that Alternative 1 recognizes the importance of grade-separating railroad junctions.
Ultimately, the EIS does not take the three good concepts – integrated service planning, component-by-component thinking, and bundling trip-time improvements when the marginal cost of doing so is low – to their full conclusion. Thus, there is no attempt at running intercity trains at high speed on shared track with commuter rail with timed overtakes, as I have proposed for both the inner New Haven Line and the Providence Line. On the contrary, the plan for capacity investment on the Providence Line includes extensive three-tracking, rather than limited, strategic four-track bypass segments. This cascades to the trip times, which are quite slow between New York and New Haven (1:08, for an average speed of 103 km/h), and a bit slower than they could be between Providence and Boston (24 minutes, whereas about 21 is possible with about zero investment into concrete).
The concepts of Alternatives 1, 2, and 3 represent bundles of levels of investment. This is the wrong approach. Alternatives 2 and 3 include new tunneled city-center stations in Baltimore and Philadelphia; but wouldn’t we want to consider city-center station tunnels in those two cities separately? It’s possible for one to turn out to be cost-effective but not the other. It’s possible for neither to be cost-effective, but for other improvements included in Alternative 2, such as curve modification around Metropark and Metuchen, to pencil out.
There’s far more interaction between different macro-level alignments, by which I mean such questions as “inland route or coastal route?” and “serve Hartford on the mainline or put it on a branch?”, than between such micro-level investments as individual curve modifications and urban tunnels. This means that instead of discrete alternatives, there should be one umbrella, taking in Alternative 2 and 3 variants, proposing all of those options as possibilities. A future study, with detailed cost figures, could then rank those options in terms of trip time saving per unit of cost, or in terms of social and financial ROI. This way, there would be concrete proposals for what a $5 billion plan, a $10 billion plan, a $20 billion plan, and so on would be.
Two elements in the study are inexcusable. First, the service plan description explicitly keeps Amtrak’s current separation of premium-fare Regionals and even-more-premium-fare Acelas. This is not how the rest of the world structures HSR: even when the HSR fares are substantially higher than the legacy rail fares, as in Spain, the fare per passenger-km is not very high, and is not targeted exclusively at business travelers. In France, the intercity fare (including TGVs, which are the bulk of French intercity traffic) was on average €0.112 per passenger-km in 2011. Premium service is provided on the same TGVs as standard service, in first-class cars. In contrast, Amtrak charges about $0.29 per passenger-km on the Regional and $0.53 on the Acela.
And second, the investment alternatives appear to include more tunneling than is necessary. I will focus on the Hartford-Providence-Boston segment in Alternative 2, since it is less sensitive to assumptions on commuter rail track-sharing than the segments overlapping the New Haven Line. It is possible to go all the way from Hartford to the western margin of the Providence built-up area without any tunneling, and without outrageous bridging; see a past post of mine on the subject here, which concludes that it’s better to just go parallel to I-95 for trip time reasons. In Providence, tunnels are unavoidable, but can still be limited to short segments, mixed with elevated routes along pre-impacted freeway corridors. When I looked at it two years ago, I saw an alignment with just 2 km of tunnel, in Providence itself. In contrast, run A in figure 9 on PDF-p. 56 says that tunnels are about 27% of new construction between Hartford and Boston, which consists of, at a minimum, about 100 km of track between Hartford and Providence.
The EIS is a step in the right direction, insofar as it does consider issues of integrated service planning and prioritizing construction based on where it can be cheaply bundled into bridge replacement. However, it fails to consider cost limitations, as seen in the excessive tunneling proposed even in areas where high-speed tracks can run entirely above ground. It’s considering more options, which is good, but, Alternative 1, while representing a golden concept, is not sufficiently developed.
What I would like to see from a study in this direction is a mixture of the following:
- Discussion of how to avoid tunnels, including various tradeoffs that have to be made (for example, above-ground construction may require more takings). Generally, I want to see much less tunneling than is currently proposed.
- A well-developed incremental option, similar to Alternative 1 but more extensive, including for example I-95 bypasses all the way from New Haven to Kingston and along strategic segments of the New Haven Line, such as in Port Chester and Greenwich.
- Greater integration with regional rail; one litmus test is whether the Providence Line is proposed to be three-tracked for long stretches, or four-tracked at a key bypass station (the options are Sharon and the Route 128-Readville segment), and another is discussion of high-acceleration electric multiple units on the Providence Line and the Penn Line.
- Unbundling of projects within each alignment – there is no need to, for example, consider the Philadelphia and Baltimore tunnels together (I also think neither is a good idea, but that’s a separate discussion). The view should be toward an optimal set of projects within each alignment, since macro-level decisions such as whether to serve Hartford are more political than micro-level ones of which curves to fix. This permits explicit discussions such as “would you be willing to spend $2 billion and slow through-trains by 9 minutes to serve Hartford?”.
Except for the first, all are kind of present in this study, but in insufficient amount for me to view it as truly a step forward. The ultimate goal must be HSR in the Northeast on a reasonable budget – closer to $10 or even $20 billion than to the Amtrak Vision’s proposed $150 billion – and this requires carefully looking at which scope is required and which is not. The EIS has elements that can be used toward that goal, but ultimately it is a step sideways, not forward or in the wrong direction.
Twenty-five billion dollars. The New York region’s political heavyweights – Andrew Cuomo, Chris Christie, Chuck Schumer, Cory Booker, Bill de Blasio – all want new Hudson tunnels, without any state funding for them; Schumer is proposing federal funding and a new interstate agency, parallel to the existing Port Authority, and a total budget of $25 billion. This is the highest figure I have seen so far; Amtrak still says $16 billion and Cuomo says $14 billion, and it’s likely the Gateway tunnels are indeed about $16 billion, while the remainder is for associated projects, such as fully four-tracking the line from Newark to the tunnel portal, a distance of about 11 kilometers. It is not my intention to criticize the cost; I’ve done that before.
Instead, I would like to point out that each time Gateway is the news, there usually seems to be a fresh cost escalation. Is it a $10 billion project? A $14 billion project? A $16 billion project? Or a $25 billion project? And what is included exactly? Amtrak does not make it clear what the various items are and how much they cost; I have not seen a single cost estimate that attempts to establish a baseline for new Hudson tunnels without the Penn Station South component, which would provide a moderate short-term boost to capacity but is not necessary for the project. The articles I’ve seen do not explain the origin of the $25 billion figure, either; it may include the tunnel and full four-tracking of Newark-New York, or it may include additional scope, for example Amtrak’s planned vertical circulation for a future (unnecessary) deep cavern for high-speed rail (see picture here).
The main issue here, the way I see it, is the interaction between public trust and political self-aggrandizement. It is common in all aspects of Israeli governance for new ministers to announce sweeping changes and reorganizations, just to remind the country that they exist and are doing something; this generally makes it harder to implement gradual reforms, and makes it completely impossible to do anything by consensus. Implementing a plan that was developed by consensus over many years makes one a bureaucrat; leaders change everything. In the US, this is the case not everywhere in government, but at least within public transportation infrastructure.
As we see in the case of Schumer’s call for a new interstate authority, the changes a heavyweight politician makes in order to appear as a leader have nothing to do with real problems that the project may have. Solving those problems requires detailed knowledge of the project at hand, which is the domain of bureaucrats and technocrats, and not of heavyweight politicians. Even a heavyweight who understands that there is a problem may not know or care about how to fix it: for example, Christie used the expression “tunnel to Macy’s basement,” invoking the deep cavern, to explain why ARC was wasteful, but chose to cancel the project rather than to remove the cavern and restore a track connection from the tunnel to Penn Station, which was in the official ARC Alt P plan until it was cut to limit the cost overruns. Managing a project is hard, and is, again, the domain of technocrats. The heavyweight will grandstand instead, regardless of whether it means canceling the project, or proposing an entirely new layer of government to build it.
As for trust, let us look at the benefits of new Hudson tunnels. The traditional, and least objectionable, is added capacity: the existing tunnels are currently at capacity during rush hour, and there’s much more demand for rail travel from New Jersey to Manhattan than they can accommodate. We can measure this benefit in terms of the combination of increased ridership from more service from more suburban areas, reduced crowding, and possibly slightly higher speeds. As a crude estimate of this benefit, current New Jersey Transit ridership at Penn Station is 87,000 per weekday in each direction. Doubling capacity means roughly doubling ridership, which would come from a combination of induced demand and diversion of traffic from cars, Port Authority buses, and commuter rail-PATH connections. This means the new tunnel can expect about 175,000 new commuter rail trips per weekday. At $10,000 per weekday trip, which is about average for very large non-US cities’ subway extensions, this justifies $1.75 billion. At $20,000, about the same as the projection for Grand Paris Express, Crossrail, and Second Avenue Subway Phase 1, all of which are justified on grounds of ridership and capacity on parallel lines, this is $3.5 billion. At $40,000, about the same as old projections for Second Avenue Subway Phase 2, which I used to analyze de Blasio’s Utica subway proposal, this is $7 billion. A $25 billion budget corresponds to a cost per rider well into the range of airport connectors.
Now, I’d like to think that informed citizens can look at these costs and benefits. At least, the fact that public transit projects only cost as much per rider as Gateway if they’re airport connectors (thus, of especial interest to the elites) or if something very wrong happened with the ridership projections, suggests that there is, normally, a ceiling to what the political system will fund. Even at $14-16 billion, the two states involved and the federal government groaned at funding Gateway, speaking to the fact that it’s not, in fact, worth this much money. In contrast, a bigger project, with bigger benefits, would be funded enthusiastically if it cost this much – for example, California already has almost this much money for high-speed rail, counting Prop 1A funds that are yet inaccessible due to the requirement of a 50/50 match from other sources.
Against this background, we see scare stories that Gateway must be built for reasons other than capacity and ridership. The old tunnels are falling apart, and Amtrak would like to shut them down one track at the time for long-term repairs. The more mundane reality is that the tunnels have higher maintenance costs than Amtrak would like since each track can only be shut down for short periods, on weekends and at night. This is buried in technical documents that don’t give the full picture, and don’t give differential costs for continuing the present regime of weekend single-tracking versus the recommended long-term closures. The given cost for Sandy-related North River Tunnel repairs is $350 million, assuming long-term closures, and it’s unlikely the present regime is billions of dollars more expensive.
I am reminded of the Tappan Zee Bridge replacement: the existing bridge has high maintenance costs due to its age and poor state, but the net present value of the maintenance cost is $2.5 billion and that of the excess maintenance cost is less, both figures well below the replacement cost. The bridge itself is structurally sound, but in popular media it is portrayed as structurally deficient. This relates to the problem of heavyweight politicians, for the Tappan Zee Bridge replacement is Cuomo’s pet project.
More fundamentally, who can trust any claim Amtrak makes about the structural soundness of tunnels? It says a lot that, when I asked on Twitter why transportation authorities do not immediately shut down unsafe pieces of infrastructure, various commenters answered “politics,” and on one (I believe James Sinclair) suggested that Amtrak order an emergency closure of one of the Hudson tunnel tracks just to drive home the point that new tunnels are necessary. I would like to stress that this is not Amtrak or a heavyweight proposing that, but the mere fact that commenters can seriously talk about it is telling. Most of the writers and commenters on the US transit blogosphere are very progressive and hate the Republicans; I have not seen a single comment recommending that the Democrats steal elections, fudge official statistics to make the party look more successful, or arrest Republican politicians on trumped-up charges, because in the US (and other first-world democracies), this is simply not done, and everyone except conspiracy theorists recognizes it. But politicizing the process of deciding which infrastructure projects are necessary for safety purposes and which are simply service expansions is normal enough that people can propose it half-seriously.
This brings me back to the issue of what I want the politicians to do, and what I expect them to do. What I want them to do is to be honest about costs and benefits, mediate between opposing interests (including different agencies that fight turf battles), and make decisions based on the best available information. This would necessarily limit costs, since, from the point of view of a member of Congress, if they get $25 billion for a piece of infrastructure then they cannot get $25 billion for another priority of theirs. They don’t do that, not in the US, and I’ve learned not to expect any better, as have the voters. Instead of working to make $25 billion go a longer way (to put things in perspective, I expect my regional rail tunnel proposal to cost $15-20 billion, at Crossrail 2 costs), Schumer is working to make $25 billion to sound like it’s going to a bigger deal than the new Hudson tunnels actually are.
None of this is a secret. American voters have learned to expect some kind of machine-greasing and politicking, to the point of losing the ability to trust either the politicians or the agencies, even in those cases when they are right. The result is that it’s possible to stretch the truth about how necessary a piece of infrastructure is, since people would believe or disbelieve it based on prior political beliefs anyway, and there is no expectation that the politicians or public authorities making those claims will have to justify them to the public in any detail. Lying to the public becomes trivially easy in this circumstance, and thus, costs can rise indefinitely, since everyone involved can pretend the benefits will rise to match them.