Future Transit Technologies

Any chance you've got any good materials on the cost breakdown on the Shinkansen line? Basically, the question I'm interested in, in this case, is what percent of the cost is the tunneling, and what percent is everything else (and, as an ancillary question, how important reducing the cost of tunneling is).
Yes, actually. However, because the railways in Japan by and large are actually private companies, a lot of detail just isn't released (or at least, not translated). Broad strokes however, because shareholders and government planning permissions exist, are released. As of 2017, the following was the expected breakdown, with construction costs not including rolling stock totaling ¥4.8 Trillion (~$43 billion US) of a ¥5.5 Trillion project.
Screenshot_20210707-142549_Drive.jpg

*However* it should be noted that this is still in the relative early stages of construction, with a crucial and difficult segment still not approved by the prefectural government. As of this April, they've announced an cost overrun of ¥1.5 trillion ($13.7 Billion USD, not accounted for above). Admittedly, they're going to the extent to underpin existing major shinkansen stations, which adds to the complexity, as does it's routing though some of Japans most varied terrain, requiring true deep bore tunnels. (Which, I should point out, completely bypasses the smaller but still substantial intermediate population centers already served by the Tokaido Shinkansen in favor of the straightest alignment between Nagoya and Tokyo, as that's already at capacity) I don't know how much of this, if any, can really be ported over to the NEC given the vast differences.

Frankly, I wouldn't be too surprised to see that soar over the next 7+ years of construction.
 
Last edited:
Yes, actually. However, because the railways in Japan by and large are actually private companies, a lot of detail just isn't released (or at least, not translated). Broad strokes however, because shareholders and government planning permissions exist, are released. As of 2017, the following was the expected breakdown, with construction costs not including rolling stock totaling ¥4.8 Trillion (~$43 billion US) of a ¥5.5 Trillion project.
View attachment 14631
*However* it should be noted that this is still in the relative early stages of construction, with a crucial and difficult segment still not approved by the prefectural government. As of this April, they've announced an cost overrun of ¥1.5 trillion ($13.7 Billion USD, not accounted for above). Admittedly, they're going to the extent to underpin existing major shinkansen stations, which adds to the complexity, as does it's routing though some of Japans most varied terrain, requiring true deep bore tunnels. (Which, I should point out, completely bypasses the smaller but still substantial intermediate population centers already served by the Tokaido Shinkansen in favor of the straightest alignment between Nagoya and Tokyo.) I don't know how much of this, if any, can really be ported over to the NEC given the vast differences.

Frankly, I wouldn't be too surprised to see that soar over the next 7+ years of construction.

Looks like 1.6 trillion yen in tunnel costs, if I’m reading that right? So, if we don’t account for overruns (which may or may not be tunnel related), we’re looking at about 33% of the construction costs and 29% of the total cost.

Very interesting. I’m not surprised that it is so large (I wouldn’t have been surprised if it was larger), but seeing it broken down like that is informative.

It looks like that could well be the one single biggest gain, reducing tunneling costs.
 
Looks like 1.6 trillion yen in tunnel costs, if I’m reading that right? So, if we don’t account for overruns (which may or may not be tunnel related), we’re looking at about 33% of the construction costs and 29% of the total cost.
Of the 1.5 Trillion, 500B is accounted for by underground challenges in excavating the station caverns in Nagoya and Shinigama. Up to you if that counts as tunnel. 300B is for the disposition of spoil and excavated material, definitely part of the tunnels. The remaining 600B is earthquake proofing the 10% of open structures.
Screenshot_20210707-142650_Acrobat for Samsung.jpg
 
Last edited:
I would certainly count underground stations as a tunneling cost, but I’m surprised at how high the earthquake measures drive up the tunnel costs.
 
I would certainly count underground stations as a tunneling cost, but I’m surprised at how high the earthquake measures drive up the tunnel costs.
If your ears weren't burning about Hyperloop overhype before factoring ^^stuff like this^^, just remember: Elon Musk wants us to believe his modal grift is a drop-in replacement for CAHSR. In California. 🤡
 
Just a quick note: the chuo shinkansen is being built, in part, because the currently-existing Tokaido shinkansen is running at capacity and there's still a market left to capture. Tokyo-Osaka tickets are incredibly expensive and part of that is because the demand is high enough that prices have risen accordingly. I almost never take the shinkansen here because bus tickets are usually 1/3 the cost (by comparison, the Hokuriku shinkansen is generally the same price as the buses). One secondary impact is that intermediate stations don't receive the service their demand suggests is necessary, notably Shizuoka and Hamamatsu.

Somewhere in the comments on Alon's site is discussion of this specific issue, since Shizuoka pref. is the one holding up maglev permits. The maglev passes through a tiny yet critical part of the prefecture, all tunneled with no stations, but the governor is claiming the geological work will destroy a critical watershed for Shizuoka's tea farming industry in a country where farmers have well-outsized political and cultural heft. The commenters all point out that, instead of harping on tea, Shizuoka should be demanding better service along its coastal mainline rail where there's latent demand.

Anyway, this is a long, roundabout way of saying that the maglev tech that the state of Maryland is pursuing is silly, because there's no latent demand between Baltimore and DC that can't be solved by the two already-existing rail lines. Electrify them both and upgrade speeds to reasonable levels and you'll cover all the demand, instead of running after dreamtech! Even a DC->NYC->Boston maglev line is ridiculous because the tunneling costs will never get repaid over the silly-low ridership (a proper Acela would cover everything for far less cost)
 
Just a quick note: the chuo shinkansen is being built, in part, because the currently-existing Tokaido shinkansen is running at capacity and there's still a market left to capture. Tokyo-Osaka tickets are incredibly expensive and part of that is because the demand is high enough that prices have risen accordingly. I almost never take the shinkansen here because bus tickets are usually 1/3 the cost (by comparison, the Hokuriku shinkansen is generally the same price as the buses). One secondary impact is that intermediate stations don't receive the service their demand suggests is necessary, notably Shizuoka and Hamamatsu.

Somewhere in the comments on Alon's site is discussion of this specific issue, since Shizuoka pref. is the one holding up maglev permits. The maglev passes through a tiny yet critical part of the prefecture, all tunneled with no stations, but the governor is claiming the geological work will destroy a critical watershed for Shizuoka's tea farming industry in a country where farmers have well-outsized political and cultural heft. The commenters all point out that, instead of harping on tea, Shizuoka should be demanding better service along its coastal mainline rail where there's latent demand.

Anyway, this is a long, roundabout way of saying that the maglev tech that the state of Maryland is pursuing is silly, because there's no latent demand between Baltimore and DC that can't be solved by the two already-existing rail lines. Electrify them both and upgrade speeds to reasonable levels and you'll cover all the demand, instead of running after dreamtech! Even a DC->NYC->Boston maglev line is ridiculous because the tunneling costs will never get repaid over the silly-low ridership (a proper Acela would cover everything for far less cost)

Well, this is a discussion of future tech, so it is not unreasonable to imagine various costs coming down. That aside, do you happen to have projected ridership?
 
Well, this is a discussion of future tech, so it is not unreasonable to imagine various costs coming down.

Problem is, the same statement can be applied to common-carrier HSR. It's still actively evolving on performance, and whatever cost savings you could eventually realize on the Maglev ROW side you can apply equally well to the common-carrier ROW and all its pre-existing scale. They're tracking in tandem, and the technological gap between modes isn't widening noticeably enough that it'll readily absorb all the infrastructure duplication. There's no 'orders-of-magnitudes' inflection point afoot like Japan is facing with its tapped-out Tokaido Shinkansen dilemma...just "replacement" mode. Mere replacement mode isn't good enough. The NEC will always have additional performance and density gears it can reach for if we're willing to pay for them, and is generations away from hitting total-saturation density where the cost of re-duplicating ALL of the right of way (including the best-performing parts) becomes palatable like it has in Japan where service density has no additional give. Yes, that might mean spending tens of billions on MOAR TUNNEL for the conventional mode to achieve a 200 MPH corridor...but if we're willing to pay that it's going to be a better deal sinking 10's-of-$B's into still-evolving conventional HSR rather than 100's-of-$B's cleanrooming everything for a largely equivalent 200 MPH maglev. You pretty much have to have a Tokaido Shinkansen-level density crunch forcing the issue, because the above-and-beyonds are so humongous for the "replacement-level" mode.

Those density conditions that are forcing the very expensive maglev complement onto the table in Japan simply aren't ever going to exist on the Acela corridor. We're in a megalopolis for sure, but not intrinsically to the degree that Greater Tokyo is a mega-mega-megalopolis. The thing that's forcing the maglev decision there is that their megalopolis is so impenetrably dense that the conventional modes can simply no longer scale to it, and service levels writ-large are starting to erode from all the mouths that have to be fed in balanced fashion. Similar conditions are not going to ever happen like that in the Northeast. Our starting density is too much lower. And, if anything, the challenges of sustaining coastal development in a climate change era are likely going to level off the density long-term. There's such cavernous service backfill left in the NEC for absorbing demand and enough Jetsons Shit innovation still to be had on the conventional HSR mode that pretty much all the transit share feasible to drive onto HSR can be well-absorbed by the common-carrier mode, if we're willing to pay for it. If novel tunneling cost-control comes into play, the highest bang-for-buck comes from steadily improving the common-carrier mode with individual segment relocations rather than re-creating the whole thing lock/stock (including max-performance straightaways) on a brand new mode. Evolution will always outpace the placement of whatever inflection point ends up calling for the maglev or hyperloop from-scratch complement...forevermore (or, at least 50-75 years out to the outermost practical limits of prediction). We're really not even close to a point of "What if'ing?" the maglev duplication. Anything hyper-expensive you can do to build that mode here you can do to straighten + densify + Jetsons Shit the incumbent NEC/Acela. Until that changes, there isn't a starting point for any form of other-modeing.
 
I think it is important to note that the Chuo Shinkansen is a parallel capacity play. The Tokaido Shinkansen isn't going away, these system will run in parallel and essentially more the double the capacity between the major endpoints.

JR Central could have done this with HSR, probably for significantly less money, but they do like sexy toys in Japan. But the key play is not the technology, it is doubling the capacity and redundancy (in case of earthquake damage to one of the lines).
 
I think it is important to note that the Chuo Shinkansen is a parallel capacity play. The Tokaido Shinkansen isn't going away, these system will run in parallel and essentially more the double the capacity between the major endpoints.

JR could have done this with HSR, probably for significantly less money, but they do like sexy toys in Japan. But the key play is not the technology, it is doubling the capacity and redundancy (in case of earthquake damage to one of the lines).
Yes...and per jlichyen's post that's the five-alarm "Where's the beef???" question behind D.C.-Baltimore Maglev in a nutshell. The parallel capacity argument is incoherent as hell because the not-at-all-slow/over-capacity D.C-BAL NEC still has so many optimization and densification gears left on the table. And yet this group is still trying to parallel it because reasons. It's other-modeing for the hawt new technological toy while leaving the box about why the parallel capacity is needed and unable to be fulfilled on the common-carrier mode...totally blank. It's hard enough trying to hang a value proposition on things so extremely expensive as this, and it doesn't help when the Hyperloopers are polluting the waters with junk science and grifting...but to come from an apparently deadly-serious starting point for study and simply dog-ate-homework the entire mission statement like that is truly amateurish and cringeworthy.

Like...even if this is wildly infeasible as a proposal it still would've been useful as all hell for DC-BAL Maglev to model exactly what service densities put a threshold of paralleling-capacity ROW on the table as a potentially future-leaning investment. We'd be able to extrapolate from there exactly how much of a service layer cake the bestest-NEC-money-could-buy would support, what the max achievable transit shares driven to that layer cake could be, quantify the redundancy angles...all kinds of enormously useful stuff for the study exercise. And they didn't do any of that. To be an involved observer from Japan who benchmarked all similar data for Chuo Shinkansen and watch the D.C.-BAL study group just space on any/all of those considerations from a similar mission-statement starting point must be...a baffling/disorienting view.
 
Just a quick note: the chuo shinkansen is being built, in part, because the currently-existing Tokaido shinkansen is running at capacity and there's still a market left to capture. Tokyo-Osaka tickets are incredibly expensive and part of that is because the demand is high enough that prices have risen accordingly. I almost never take the shinkansen here because bus tickets are usually 1/3 the cost (by comparison, the Hokuriku shinkansen is generally the same price as the buses). One secondary impact is that intermediate stations don't receive the service their demand suggests is necessary, notably Shizuoka and Hamamatsu.

Somewhere in the comments on Alon's site is discussion of this specific issue, since Shizuoka pref. is the one holding up maglev permits. The maglev passes through a tiny yet critical part of the prefecture, all tunneled with no stations, but the governor is claiming the geological work will destroy a critical watershed for Shizuoka's tea farming industry in a country where farmers have well-outsized political and cultural heft. The commenters all point out that, instead of harping on tea, Shizuoka should be demanding better service along its coastal mainline rail where there's latent demand.
Since you have local insight, and as I haven't been in 5 years now, do you see the chōū shinkansen basically supplanting the Nozomi limited stop service, with it's slots going to the Hikari/Kodama services, or do you see the Nozomi continuing to run with its current frequency?

Also, I assume some part of this must be mode diversion from Air. What amount of airline traffic between these cities hasn't already been captured, and is the Chōū Shinkansen likely to capture the rest of it? There still is a huge number of flights between these 3 core cities.
 
I'd like to describe a form of travel I can manage here which I've never managed in the US. When I've gone on business trips I'll sometimes hang around wherever I've visited, sometimes visit a park or museum before catching the shinkansen back home. I'm able to do this because all shinkansen lines run 4~5 (or more) trains per hour, so if I "miss" my train I can simply catch the next one. This is usually very easy because I have an unreserved ticket which I can use on any train (like MBTA commuter rail). Back when I lived in the US I had one or two occasions where I had to cancel weekend plans because I missed my NEC regional train and the next one would mean not arriving at my destination in time.

EDIT: what would make the NEC look like tokaido is if you could ride all super-local regional trains from Richmond VA to Portland ME, and at no point did mid-day service dip below 2 trains/hour (on the outer ends, I expect better for core areas). Meanwhile, NEC regionals function as Acela "locals" running Acela trainsets at Acela speeds, likely on 90% exclusive tracks (ex: four-tracked the whole route, six-tracked around NYC, NSRL an obvious pre-req). Once you have that network and it's saturated beyond the point of maximum service, then (and only then) should you be discussing maglev.

Well, this is a discussion of future tech, so it is not unreasonable to imagine various costs coming down. That aside, do you happen to have projected ridership?
Unfortunately, I don't, and the Ministry of Land Transport's website is extremely unhelpful because their site is filled with meeting records with the obstinate Shizuoka pref. officials. This document, however, does state on slide 11 that the max capacity of the maglev as planned is 10 trains/hour, holding 1,000 pax/train, meaning 10,000 pax/hour - compared with the regular shinkansen which has a max capacity of 17,000 pax/hour. Also note on slide 13 a visual comparison of the size between existing shinkansen cars vs. new trains. (nbd: the file is from 2010, don't hold it against me if it's already out of date)

But this also emphasizes that the maglev is a supplemental service, not a replacement service. To answer Sitlin's question, my guess is that Nozomis would not run with current frequency (since the point is to open up frequencies to Shizuoka stations) but I can't say with certainty yet because the maglev skips Kyoto, a key Nozomi stop.

As for airlines, it really depends on how cheap regular shinkansen tickets plunge. A roundtrip plane ticket is still less than a one-way shinkansen ticket, so there's definitely still going to be a market for ultra-cheap travel.
 
Unfortunately, I don't, and the Ministry of Land Transport's website is extremely unhelpful because their site is filled with meeting records with the obstinate Shizuoka pref. officials. This document, however, does state on slide 11 that the max capacity of the maglev as planned is 10 trains/hour, holding 1,000 pax/train, meaning 10,000 pax/hour - compared with the regular shinkansen which has a max capacity of 17,000 pax/hour. Also note on slide 13 a visual comparison of the size between existing shinkansen cars vs. new trains. (nbd: the file is from 2010, don't hold it against me if it's already out of date)

But this also emphasizes that the maglev is a supplemental service, not a replacement service. To answer Sitlin's question, my guess is that Nozomis would not run with current frequency (since the point is to open up frequencies to Shizuoka stations) but I can't say with certainty yet because the maglev skips Kyoto, a key Nozomi stop.

As for airlines, it really depends on how cheap regular shinkansen tickets plunge. A roundtrip plane ticket is still less than a one-way shinkansen ticket, so there's definitely still going to be a market for ultra-cheap travel.

I meant for the theoretical US maglev line.
 
I meant for the theoretical US maglev line.
Basically, all of your questions for that thing can be answered by digging through their DEIS website and documents posted there. Their ridership study is heavily redacted, but this is the core line that wasn't.

Ridership following the end of the ramp up period grows from approximately 16.3 million annual trips in 2027 (45,000 daily), to approximately 24.5 million annual trips (67,000 daily) at the model’s forecast horizon of 2050.

That's.... An incredibly ambitious number, which I consider more or less fiction. I have massive questions about their methodolgy and peer review process (also all redacted), as they must be using some ridiculously rosy assumptions. For context, the entire NEC in FY2019 Acela and Regional combined did 12.3 million (~33k daily) in ridership, including 8m originating or departing NYP, and only 900k @ Baltimore.

Admittedly, the MARC Penn Line does 24k daily. To get to their numbers, you'd basically have to assume that every single passenger stops taking Amtrak and MARC and changes to this thing - then almost double it. I don't see it happening, with only 3 total stations, 2 being urban terminals and an airport station? No way. I'm not even convinced that there's that many people going between those 2 cities every day

Even with their numbers, that's a volume of people that's readily absorbable by just improving existing NEC services between MARC and Amtrak, if we take even 5k pax/ hour as max throughput on the NEC. Also, as far s travel times go, between Baltimore and DC, MARC is scheduled for 64 minutes, Regionals 47, Acela 33, which will come down with faster trains and improved infrastructure. This maglev is proposing 15. That is not a big enough difference to justify it's existence.
 
Last edited:

Back
Top