Paris, World Capital of Expensive Regional Trains

I have found something a European city does worse than the United States in public transit. Paris has just announced its new bilevel design for the RER B, currently the only line running single-deck trains due to restricted clearances. The new double-deckers, dubbed MI 20, are expected to cost 2.56b€ for 146 trainsets, each 104 meters long, for a total of 168,600€ per meter of train length.

I’ve criticized Paris’s use of double-deckers in the past. The cost premium for a double-decker, usually around 25-50%, at best matches the gain in seated capacity, and leads to other capacity problems with access and egress, which are of especial importance on urban rail like the RER. Not for nothing, bilevel trains are not used in Tokyo except for the occasional first-class car (“green car”), which is less crowded by design than the legendarily crowded subway and regional rail cars.

However, this is a lot worse than the usual premium. The only comparably expensive bilevel I can find is the Stadler KISS order for Caltrain, which at $230,000/m for the base order (and only $160,000/m for an equal-size option) comes at a large premium over usual KISSes (both around 130,000€/m) due to client interference and micromanagement coming from low competence by American railroaders.

But the KISS is a high-performance train, at the expensive end in Europe, too. Moreover, it is fully bilevel, whereas the MI 20 has a mix of single- and double-deck cars, with high-platform boarding. Comparable split-level trains go well below 130,000€/m. Canalblog has a compendium of recent Coradias: the single-level example for Milan is 6.25m€ per 84-meter train, or 74,400€/m, and the mixed single- and double-deck examples are 96,700€/m in Luxembourg and 117,600€/m in Germany. The mixed-deck Siemens Desiro HC has a range of costs: its RRX order is 1.7b€ for 82 150-meter trainsets, which is 138,200€/m, but a smaller order for the Berlin RegionalBahn is 300m€ for 21 six-car and 2 four-car trainsets, or 89,600€/m, which is a high but not unheard of cost for a single-decker, let alone a double-decker. The Desiro HC is being delivered to Israel as well, at a cost of 900m€ for 60 trainsets totaling 330 cars, or 109,100€/m.

There’s nothing special about Paris that justifies such a cost – the highest in the world so far, even beating the Americans. Rather, the problem is most likely that Paris thinks it’s special and won’t buy a standard platform. Canalblog points out that the Coradia Duplex formed the basis of the X’Trapolis, currently delivered for the RER D and E – and the X’Trapolis’s first tranche spent 29% of its budget on design and engineering, driving the cost up to a stratospheric 21.83m€ per train of length 112 or 130 meters. Even averaged over the entire order of 255 trainsets, at which point economies of scale kick in and the bespoke design is less harmful, the cost is 121,400€/m, which is 25% more than the more standard Luxembourg design.

Update: Clem Tillier asked me about Madrid’s recent Cercanías order. This is a mix of Stadler trains and Coradias, both mixed single- and double-deck; the Coradias, using the same platform as the Paris X’Trapolis trains and built in Spain rather than in France, cost 1.447b€/152 trainsets, or 95,200€/m, and the Stadlers, mixing KISS and FLIRT technology, cost 998m€ for 24 100-meter trains and 35 200-meter trains, or 106,200€/m.

In a megacity like Paris, it’s tempting to think one is special and must have special equipment. But the resulting high costs are particularly damaging in such a city. The RER B runs every 3 minutes at rush hour, which means that high rolling stock costs are proportionally a bigger problem than on a less frequent system. The cost premium of the order over standard single-deck trains is a factor of around 2; half the cost is 1.3b€, which would be enough to build some necessary tunnel extensions, like quad-tracking the combined two-track tunnel for the RER B and D between Châtelet-Les Halles and Gare du Nord, or if RER investment is not desired then around 6 km of tunnel for Grand Paris Express after the latest cost overruns.

France needs to let go of its pride and recognize that Paris is merely the largest city in the Union, with the same standards and regulations as the other 440 million of us who do not live in Ile-de-France. Vanilla Coradias and Desiros that work elsewhere should also work for the RER, with minor tweaks to take into account high platforms and the loading gauge, both of which the vendors are experienced in dealing with due to common intra-European variation. The people who sign extravagant contracts may feel special about the train design, but the passengers who end up not getting the investment the cost premium would have gone to are going to keep feeling packed on rush hour RERs. The region ought to do better and hire managers who are better than this.

163 comments

  1. Tom the first and best

    If they are spending double-decker money, they should do so efficiently. Double-decker trains would be much more efficient with double-decker platforms, which are presumably significantly more achievable these days with effective platform screen doors* and in-cab CCTV platform monitoring. Remodelling the stations to have double-decker platforms drives the cost too high compared with the alternatives.

    * Which are a must have anyway for platform safety and particulate pollution control (rail tunnels have lots of particulate pollution, especially those of the RER).

    • Herbert

      Is there any place in the world with existing double decker platforms?

      Nothing against innovation, but methinks such a delicate operation should be trialled first in a place where “it just not working” – or even just the inevitable teething issues – wouldn’t have as high a cascading negative effect as on one of the world’s busiest regional urban rail systems…

      • Tom the first and best

        I am not aware of any current or previous double decker platform systems and am fairly sure there aren`t any.

        Most of the technology has existed for them since the 1960s. Platform screen doors are basically lift door for horizontal lifts lifts and the automatic train operation needed to get door alignment has been used since the 1960s (on the Victoria line). CCTV monitoring, although the required screens may have had a size issues in the days of CRT screens.

        Such a system would be easier built in a open air railway as underground station reconstruction would be more expensive.

        • Herbert

          Not to mention subway systems do not usually have the height clearances required

          • Eric2

            If we’re already talking bilevel trains, the tunnels are high enough.

          • Herbert

            Which subway has bilevel trains?

            I know of bilevel trams in Alexandria or Hong Kong or Blackpool, but bilevel subways? Never heard of any…

      • Eric2

        I suspect the complications involved would be minimal. If you can line a train up to the lower platform, then it will be automatically lined up to the upper platform.

        • Herbert

          We test systems in the real world because no plan ever survives first contact with reality

  2. helm

    To add: The expensive RRX order contains 32 years of maintenance.

    I think Paris is held kind of hostage by Alstom. They couldn’t order a train that is not produced in France, Paris showing it gives jobs to the province. So Alstom gains every order and can extract as much value as possible. All the recently ordered RER, Transilien, Metro and Tram trainsets come from Alstom. Do they have also cost premiums?

        • Mikel

          The process has been litigious, since Alstom is definitely salty about the erosion of its monoploy over the French market. The CAF-Bombardier consortium was initially awarded the RER contract, but then Alstom bought Bombardier and withdrew it from the competition, thus throwing CAF under the bus (ha) and leaving Alstom-proper as the only offer standing. Then RATP threatened to sue Alstom for anti-competitive behaviour and forced them to keep the original joint offer with CAF. So in the end Alstom got half of a big contract and a completely avoidable PR disaster.

          The story is also complicated by the side plot where the EC forced Alstom to sell its Alsace factory as a condition to greenlight the merger with Bombardier. The two possible buyers were Skoda (an option preferred by Alstom’s leadership) and CAF (supported by the unions). In the end CAF withdrew for reasons that are unclear to me. I wonder if all those corporate shenanigans had some effect on the final price of the trains, since Alstom complained that CAF and Bombardier had underbid.

      • Onux

        Bombardier is still Francophone, however, originally founded and headquartered in Quebec. I don’t know enough to know if that counts for something in France, but it might.

        • Jojo

          Well, no. It was Canadian and Canada didn’t belong to France for a couple centuries.
          However Bombardier Transport was headquartered in Berlin.

          • Herbert

            But Bombardier had production facilities in most European countries…

    • michaelrjames

      Well, the train was ordered from Bombardier long before Alstom bought them.

  3. michaelrjames

    I’m a bit confused on this, and am wondering if you can clarify. Because earlier this year Alstom was making contrary claims, as per the extract from Rail Journal below. Your report suggests that Alstom has been successful in forcing a renegotiation, in the courts, of the original Bombardier contract upwards since Alstom took over their train division.

    https://www.railjournal.com/fleet/alstom-calls-for-e2-56bn-rer-line-b-fleet-contract-to-be-renegotiated/
    Alstom calls for €2.56bn RER Line B fleet contract to be renegotiated
    David Burroughs, 05 Feb 2021.

    After it completed the acquisition of Bombardier on January 29, Alstom reviewed the bid submitted by the consortium before announcing on February 3 that it planned to attempt to renegotiate the contract. Alstom chairman and CEO, Mr Henri Poupart-Lafarge, said there were concerns that the price offered for the contract was too low, and there was not enough flexibility within the technical specifications.
    “We examined the offer very carefully from a technical and financial point of view and came to the conclusion that it was not viable,” Poupart-Lafarge told French newspaper Le Figaro.
    Poupart-Lafarge said there was also a large gap in the cost of the bid submitted by Alstom and the Bombardier-CAF offer, which he said was “too good to be true.” “There is no margin for error in the technical choices, the equipment choices are made as precisely as possible, and the gauge of the trains is not up to standard,” he says. “The development schedule for the trains is 50% shorter and the price 20% lower than those of the new generation RER E, on which Bombardier is losing money.”

    It is also not clear how many of the other duplex trains you cite could function on the line as parts of it (presumably the old southern tunneled part?) apparently has more restrictive loading gauge issues. Without this problem IdFM (Île-de-France Mobility) would have long ago adopted one of the duplex trains giving perfectly good service on the other RER lines?

    Just for clarity–on issues we have discussed before–you say that MI20 will be used for peak 3 min headways, ie. 20 tph. Only when heavily loaded trains like the 2,600 pax RER-A duplexes approach 30 tph are there any egress issues that induced a slight reduction in frequency to 24-28 tph; and that this only applies to the short, intensive peak period. And which will disappear shortly as the M14 northern extension and RER-E western extension/takeover of a RER-A branch, relieve pressures on RER-A (busiest commuter line in the world outside Asia). Some of the many adaptations and evolutions of Paris’ transit system are in reaction, at least in part, to avoid the extreme crowding issues as seen in Tokyo but unacceptable in the west. While you may be correct that these trains are ‘too expensive’, one has to put it into this wider context. Not only does it appear to be a court-mandated cost increase–and thus not a discretionary or notional ‘specialness’ of Paris by IdFM–but is there a western mega-city transit system as good as Paris? London’s first RER line (Crossrail/Elizabeth Line) is costing US$26bn and counting, is running 3 years late (actually about half a century late), and shows what the alternative can be: vastly more expensive relative than if built when first conceived (actually in the late 1940s). Perhaps Paris’ willingness to pay a bit more today is more prudent–economically and operationally–than the Anglosphere habit of postponing such capital costs forever? Oh, and let’s not forget the other metrics of note: IdFM carries twice London Underground and average ticket prices are at least twofold less expensive. My point is that, compared to its peer cities of London and NYC, as a (former) Parisian I would be willing to give those who make these decisions for Paris transit the benefit of the doubt. In all the really important metrics they get things right, if never perfect.

    Also–and we’ve had this discussion before too–the use of “cost-per-metre” is very uninformative, indeed quite misleading. I’m sure it is one of those historical bits of baggage but for me that doesn’t make it valid. It’s like coming across some of those ancient Imperial measures (like stones, quarts, or fluid ounces etc) only worse, because at least one could find a calculator/converter for those things. I can’t do that with this metric and cannot possibly be bothered to seek out a metric that would be appropriate for such a comparison: cost per pax carried (and for both seated and total, seated + standing).

    • Herbert

      Pax carried on urban rail correlate heavily with linear length. You can (and in some cases people have) remove seats from a train or move them around and all you get when increasing seating capacity on a train of given length is reducing legroom…

      As per standing capacity, there are supposed to be rules as to how many standees per square meter, but of course those aren’t really enforceable if the literal push comes to the literal shove… And then there is the issue of what to do when higher distances between pax are desirable, such as during a global public health crisis…

      Of course linear meters are not exactly equal to floor space, but on urban rail width (well, external width anyway) is basically a fixed quantity for any given system as the loading gauge cannot (easily) be changed and using narrower trains on wider lines (Google “Berlin U-Bahn Blumenbretter”) is only ever done as a stopgap in extremis…

      • michaelrjames

        @Herbert: “Pax carried on urban rail correlate heavily with linear length. ”

        That only applies to single-level trains, and even there the wide gauge trains that allow 3+3 provide exactly 50% more seating. The ‘new’ Velaro e320 Eurostars carry 25% more pax than the old Alstom TGVs only because they have 3+2 seating and they can only do this because they are 300mm wider trains, which cannot be used on most of the French rail network. Part of the reason why SNCF wanted duplex TGVs.

        It clearly doesn’t apply to duplex trains which Alon explicitly and implicitly admits: “The cost premium for a double-decker, usually around 25-50%, at best matches the gain in seated capacity … ” So this means that if you used a metric I suggested (per pax rather than linear metre of train) then the duplex RERs cost would be almost the same. The duplex MI-09 trains on RER-A have 948 seats while the earlier MI-84 (same trains on current RER-B) have 624 seats, exactly 50% more seated pax per train. MI09 trains carry 2600 seated + standing pax. That is, it’s not really a ‘premium’ at all! The extra cost buys exactly the extra number of seats you’d expect. It takes a particular kind of econocratic mind to say the cost of duplex trains is 40% or whatever higher than simplex with no improvement! The ‘problems’ of egress etc are really an exaggeration, only at above 25tph and only at extreme peak hours, and only applicable in Asian megacities with extreme populations. That’s an awful lot of conditionality to work oneself into indignation on costs! And as I said, is being solved by the more sensible option of sharing that huge pax load of a multi-branched line spread over about 100 km east-to-west across a mega-city. The M14 northern extension has presumably already brought some relief to RER-A, though the biggest relief will come from RER-E taking over one of the RER-A western branches.

        Perhaps ‘linear metre’ might work if the duplex sections were counted, ie in addition to the carriage length add in the length of the upper level of the carriage … at least then you’d be comparing apples to apples instead of to oranges. Though it is not much of a metric when it doesn’t reflect accurately the 25-50% difference one can get with even single-level trains depending on width or seating configuration. And even then … I once travelled on a 3+3 duplex commuter train between Shanghai and Hangzhou which must have carried a humungous number of pax fully loaded (which it was), much more than a fully-loaded RER-A. It was fine but then I’m a smallish slim person and doubt that an average American would have thought the same! Those trains would have probably been configured 3+2 for American use.

        • Herbert

          Double decker trains have obvious disadvantages when it comes to quickly loading and unloading, which is why they are virtually never used on S-Bahn type services – certainly not on those as busy as Paris RER.

          Also: seated capacity may be a relevant measure for long distance trains (for which loading gauge issues are a bit different as national or even international networks tend to have wider variation of loading gauges than urban networks) but in an application where virtuality nobody is on the train longer than an hour (and most considerably shorter durations of time) it becomes much more desirable to offer standing room so people can get on rather than trying for a “premium” product for a select few with “seating only”…

          And if you have width (well, exterior width, but then there’s only so much finagling more interior width with given exterior width) a given due to loading gauge issues (for any given system it makes no sense to build trains to either a smaller or a bigger loading gauge than what your network is designed for) than the only measure you can vary is length – and of course the maximum length of each given train is determined by platform length…

          • michaelrjames

            @Herbert

            No, no and yes.

            I mean, all this fuss about a problem that only happens occasionally and only if trains are run at greater than 25tph and fully-loaded (2,600+ per train!), ie. during a short peak hour. You’ve never heard of these problems on RER-C which was running duplex trains long before the other lines, because they don’t have the traffic and thus the need for such a frequency etc. Or for that matter the Sydney commuter trains, ditto. It simply isn’t the issue some claim it to be.

            Seating capacity is definitely a worthwhile issue on the RER network. Even on 20km or really >5km journeys let alone the 50km journeys some do on the Paris RER. Or think about those stressed Brits paying £4,000+ pa for a season ticket and travelling from, say Brighton or anywhere on the southern coast, about an hour to London and being expected to stand! (It happens and should have caused riots. It did cause lawsuits which pax lost.) RER duplex trains have a tonne of standing capacity (in vast foyers which also have strapontin seating) as well as the high seating capacity, and by definition tend only to be fully-loaded in the last km into Paris. It’s a better solution than Crossrail (if it ever gets into service) with its exclusively longitudinal seating and more people forced to strap hang on its very long journeys. Crossrail is modelled on Paris RER-A but they haven’t taken this pax comfort aspect into consideration … which is the recent half-century history of the hapless British rail traveller. Incidentally I am quite sure Alon expected to sit in comfort on his RER journey to IHG in the Paris hinterlands …
            which I absolutely endorse.

            On the third issue, of course you can’t build what the network can’t handle except where a large-scale reconstruction of the network is considered, or built de novo (like China in the modern era). That’s why the French went for duplex TGVs (after doubling train length wasn’t enough), while the Germans can apparently handle the extrawide trains, and the British system can’t handle either so are stuck with a system that is hard to improve, especially if the state doesn’t want to spend money especially on capital projects; the dramas over HS1, HS2 and Crossrail proves this.

          • Herbert

            If you add even just fifteen seconds to every stop due to access and egress issues, you’ll have ten minutes lost by the time you reach forty stations…

            And we all know that even “small” time savings can be crucial in an ITF. Again, there’s a reason virtually no S-Bahn – and no subway – has double decker trains…

          • Herbert

            Continental Europe agreed on “Berne gauge” at some Belle Epoque conference (British loading gauges were even more constrained, because penny pinching – exceptions included the GWR which was initially built to much more expensive standards, including Brunel gauge)… Basically it was a compromise to slightly raise French standards while not bringing everyone to German or Nordic standards, which were higher.

            Of course if you build hsr – especially if you do it like Spain with several deliberate incompatibilities – you could raise standards… But for whatever reason France seems to have refrained from doing so…

            Of course the tgv duplex is less of an issue, because the vast majority of tgv pax only ride one stop. If you have a takt system like Germany with numerous intermediate stops, access and egress becomes more of a pressing concern (and even then ICE schedules sometimes print separate arrival and departure times on official schedules which should tell you all you need to know about dwell times…)

          • Herbert

            Also: a public transit system that doesn’t work during the highest demand hour of the day does not work, period.

            That doesn’t mean it has to work only at that time, but “ah, don’t worry about it, it’s only an issue when everyone wants to ride” ain’t gonna fly…

          • Herbert

            A second class Bahn Card 100 costs roughly four thousand euro. It is of course valid on all trains, not just the lowest quality commuter trains…

          • Herbert

            If you’re going 20 kilometers or more, shouldn’t (part of) your trip be on a more “express” type of train?

        • Jojo

          Wat ? No, of course Velaro e320 Eurostars don’t have 3+2 seating, and they are only about 2,92 m wide. You’re mixing England with China.

    • Alon Levy

      A couple points.

      1. Cost per meter is pretty consistent within each class (single- or double-deck). All the single-deckers and LRVs cluster in the same area, and then the Americans pay a premium because of political micromanagement.

      2. At no point am I comparing anything to the UK, which IIRC has always had a noticeable rolling stock cost premium over the Continent. We’re one Union, even Germany is starting to engage in pro-Union foreign policy now that the Anglos are conspiring to screw over France with the submarine deal.

      3. The RER B needs fast egress even more than the RER A: 20 vs. 30 tph helps a lot, but on the other hand, the RER B’s morning rush alightings are concentrated at Les Halles, whereas the RER A’s are spread between Les Halles, Auber, Etoile, and La Défense. The rush hour dwells I observed were around 60 s even with single-deckers and 32 doors per ~200 m train.

      • michaelrjames

        1. If true then why do you use it to compare across classes? And anyway, is 25-50% variation = “pretty consistent”? C’mon, admit that it’s a crap metric.

        2. Fair enough though I don’t think a bit of a premium would matter if the service was good enough. It ain’t. (For clarity, in Paris it is … good enough, or better than good enough.)
        Unfortunately screwing over the French was not the object of the conspiracy. It was entirely an afterthought. Or a thoughtless after-effect. I am in the crowd who believe it is one of the worst strategic decisions w.r.t. Australia and Asia, ever. Though current Labor is totally supine and is going along with it (because the belief is they lose on any ‘national security’ issue no matter how loony or irresponsible the concept, and this one is massive and irreversible). At least two former Labor PMs, Rudd & Keating gave the right response, and the essence of it is what Paul Keating said about 3 decades ago: Australia had no choice but to stop looking for our security from Asia and start looking for it in Asia. This dumb action manages to alienate most of Asia, including China, superglue us as deputy-sheriff to the fading powers as well as alienate the only European nation that has several million citizens in the India-Pacific. Oh, and we will be lucky to get a submarine by 2045, which compares to the complaint about it being early 2030s for a French one … but really the submarines are the least of it.

        3. Good try but I don’t buy it. Duplexes will fix the line up for another half century. Though by then they might have doubled that tunnel pinchpoint …

        • Herbert

          If you’re doing 30 trains per hour, trains are – at most – 120 seconds apart. If a train is stationary for (at least) 60 seconds in a station while the next keeps moving towards it, that reduces the temporal separation you can actually work with to 60 seconds… Not ideal…

          Add extra time by it being a double decker and you drastically hamper reliability…

  4. Phake Nick

    What do you think about Tokyo mayor’s double decker train suggestion, which include two layers of platform and thus minimal need for in-train staircase, maximizing capacity compares to current double decker train in operation, in addition to improved ingress/egress time comparable to ordinary trains due to top and bottom layer passenger flow do not interfere?

    • michaelrjames

      @Phake Nick

      It sounds terrific and Tokyo certainly needs something. Other than cost, the main problem, I am guessing, is the massive disruption in getting to such a system from today’s. Is it at all practically possible? Reminiscent of south-east England which assuredly rail experts knew in the immediate-postwar period should have been upgraded to European freight gauge but weren’t, leading to innumerable problems and massive costs in alternatives (eg. HS1, HS2, 12 years of compromise in running Eurostar).
      I think the question applies even more to China’s megacities. They have built them (or hugely expanded them) very recently, ie. with full knowledge of the problems such megacities like Tokyo create. Building so much virgin system they had the opportunity to do this (and quad-tracking for express/local trains on their often very long lines) but haven’t.

      • Herbert

        As above, such a system should be tested on some lower profile line before you do open heart surgery on the urban rail system of a major world city like Tokyo or Paris…

      • Phake Nick

        I think it is definitely beyond their expectation to have such big influx of population into those cities, with the fact that Hukou system still exists in Japan fixating one’s right to their place of origin, despite that’s how those cities succeeded and have the factories developed.

    • Sassy

      There’s a lot lower hanging fruit than double deck trains in Tokyo.

      CBTC moving block could reduce headways and increase frequency. Very few lines in Tokyo are even at 30TPH, and going from 30TPH to 40TPH is literally a comparable capacity gain as double deck trains, and improves waiting times. Though realistically, many lines would use any gains to just run more limited stop services.

      If you want to pour concrete, then some of the most crowded lines aren’t even running full size trains yet, such as literal people movers such as on Nippori-Toneri Liner, or just trains with fewer than 10 cars such as on the Nambu Line, and could be upgraded to full size 10+ car trains without something exotic like double deck platforms.

      • Herbert

        Is forty trains per hour a feasible stable schedule?

        That’s a nominal 90 second headway and of course you need to include a fudge factor (five seconds? Ten?) Of trains bunching slightly for whatever reason – if trains are fine doing exactly 90 second headways in a vacuum but not a 85 second headway under daily operational conditions, then you shouldn’t be writing schedules based on that…

        And yes, I know Nuremberg U-Bahn does nominal 100 second headways with interlining and ATO but their system has flaws, too, as anybody who’s ever been on a U2 or U3 that has a computer malfunction and just… Stops (until a person from the public transit company comes and moves the train to the next station in manual mode…) can attest to…

        Apparently besides the high cost for questionable benefit and the potential issues with the union (when U2 was automated all former drivers were redeployed to customer service roles and U3 having never been driver operated didn’t have the issue) a reason why U1 is not getting automated any time soon is that the “emergency brake if something is on the track” is (had to be?) set so “trigger happy” that while it is working fine on the (almost?) fully underground U2&U3, it would have issues on the largely above-ground U1 when stuff like tree branches, various detritus or even leaves falls on the tracks…

        • Herbert

          Furthermore, I think keeping enough staff current on how to manually operate the trains (as you’ll always need, per above) is probably a lot more difficult if the only manual operation is during emergencies or training… Wonder how the Rennes metro does it…

          • Phake Nick

            Some systems in Japan, like Yurikamome, do have their system fully automated. They are also a separate company so they cannot just draw train driver from elsewhere during emergency. I think they do let their staffs manually drive their trains every once in a while in order to retain the familiarity with the driving system.

          • Herbert

            Even if it were legally not an issue for – for example – Nuremberg U-Bahn to rely on train operators from Munich (they lent each other rolling stock back in the 1970s) it would entirely jeopardize the ability to respond rapidly to issues. If you have a train full of pax stopped between stations for no reason that passengers can discern, you want the reaction time measured in minutes not hours – even if you can phone in soothing messages from the central dispatching office…

            So, in essence, you need someone on standby to do it as needed and you need them in geographic proximity… And at that point, it might as well be an employee of yours on a break that pays better than normal breaks or someone on a (shudders) split shift…

          • Phake Nick

            Last time Yamamote line stopped working due to electric failure, it took them at least one or two hours to start letting trapped passengers out of the trains and walk to the nearby station. And a few winter ago, a train in Niigata got trapped between stations due to weather, passengers have to stay in the train for almost a day.
            In situation where you want to get out of a train as quick as possible like when there are fire, either due to mechanical failure, accident like passenger belonging battery fire, or arson, or when there are someone actively trying to endanger the passengers with tools, ranging from insane passenger with knife to full on terror attack, it is quicker and safer to evacuate passengers by driving the train to the nearest train station, open the train door as usual, and let passengers leave the train through regular door on regular platform, than trying to evacuate people right at the spot between the station, given that most commuter trains or metro should be no more than ~5 minutes away from their nearest station at any given point along the line and it would only take some seconds for passengers to all get out of a train when a train stop at platform as usual. On the other hand, if you immediately stop at the track, for metro lines the passengers can only exit through a train’s front door and back door in driver’s cab, while for commuter rail line it is necessary to first ensure there are no oncoming traffic at the neighboring track before passengers can be allowed to leave the train, both of which would take far more time than just stop at the next station
            Unless, somehow your train is trapped in the middle of flooding, or the line is down after earthquake and tsunami warning have been issued and your train is now stopping near the coast, then emergency evacuation will be needed while the train cannot operate and cannot wait till situation clear, then in ordinary Japanese trains, there are evacuation staircase and evacuation slider to help passengers leaving the train, with the fact that most commuter trains are high-floor in mind. In such sort of situation, emergency sliders and staircase can also be prepared for people on the upper deck, so I don’t think there would be differences.

          • Herbert

            Emergency scenarios need to account for the worst possible scenario – i.e. For whatever reason the train can’t move on its own power and passengers can’t stay on the train – for good measure add that whatever electric doohickey usually opens the doors not working, forcing you back to whatever manual fallback there is (is there one? There better be…) oh and it’s pitch black darkness, because why not.

            So, what then? Your plan better not be *shrugs* “guess I’ll die, then”

          • Herbert

            So it would appear you do not have a plan for that case… Do I get that right?

        • Alon Levy

          M14 and M1 in Paris have an actual 85 s rush hour headway, but that’s in context of full segregation. Tokyo doesn’t have that except on the Ginza, Marunouchi, and Oedo lines, and the first two are IIRC the highest-frequency lines in the city.

          Th brutal response to all of this is to deinterline Tokyo as much as possible – for example, end the Chuo-Sobu with Tozai interlining, to create a self-contained line from Nakano to the Toyo-Katsutadai. But the trend is going the other way due to competition between railways.

          • Sassy

            Competition is just responding to what people want. Is it hard to believe that people might prefer trains that take them to their destination faster and with fewer transfers but are crowded, over less convenient less crowded trains?

            If people really wanted a slower trip with more transfers but a higher chance at getting a seat, then a company could switch to running primarily transverse seating trains, and cut limited stop and through run services, and make a ton of money. But they don’t.

            Is it bad that trains run the way people demand they be run?

      • michaelrjames

        Right. So you’re saying that the über-crowding is company policy. Sounds entirely credible … in a for-profit system. In the pampered west, anything approaching those levels of discomfort is not tolerated and provokes solutions that involve spending (public) money on new signalling, new trains and new lines. Well, except the privatised British system where the busiest franchise operation, the Southern network in SE England could definitely have benefited from running duplex commuter trains (from the Sussex & Kent coast etc to London) because they don’t involve more trains or more staff, ie. more operational cost. They actually reduced the number of trains so as to produce a better on-time record (but worse pax outcome). Of course the lines cannot support duplex trains.

        • Herbert

          I’m pretty sure the train operating companies don’t get to set schedules… So not having more trains is on whoever is writing the tender, isn’t it?

        • Phake Nick

          The crowdedness on Tokyo commuter lines were worse when Japan’s National Railway was still under the government
          Also the question is how crowded is too crowded. When crowdedness in trains commuting to Tokyo was up to 300%, it did caused them to spend a lot of money building many new lines, but now it is 180% crowded, is it still too much or should further investment be spent to build new lines so that the crowdness is down to 100% even during peak time? Thanks to the coronavirus pandemic we can now see the situation where trains crowdedness are down to 100% in Nagoya and Osaka even during commute hours as revealed by 2020 Japanese government report, but given the resultant balance sheet of those companies in year 2020, is that sustainable even if one discount the cost to build those lines? As even when one assume those lines cost nothing to build, it is still necessary to operate and maintain the trains, and is a commuter line unable to provide sufficient demand packing the train up to line capacity limit cost-efficient to continue operating?

          • Alon Levy

            The crowdedness in Tokyo is a matter of fiscal austerity (the state demands a 30-year payback period, which is less than what’s available to the Japanese private sector) and high city center construction costs. A Tokyo with only the austerity practiced in Germany, let alone France, would build more subways, as would a Tokyo with the construction costs that Paris had until ~5 years ago.

          • Sassy

            The crowdedness in Tokyo is also fairly intentional. 180% is 100%. When crowdedness gets low enough, more limited stop services and through running gets added, since a faster more crowded ride is preferable to a slower less crowded ride.

            Which is why I suspect CBTC and moving block would not reduce crowding, even though it could. Instead of reducing crowding, it could be used to improve convenience.

          • Herbert

            Don’t worry, Japanese demographics will make the problem disappear soon enough…

          • Phake Nick

            Japanese demography is aging and shrinking but they also increase concentration into large cities. If I recall correctly according to the current trend, Central part of Tokyo will still see population grow at least until 2060s

          • Herbert

            Do people in Japan still crowd into Tokyo past retirement age? And do retired Japanese people ride public transit much during peak hours?

          • Phake Nick

            I cannot tell for sure exactly how much there are (I am sure there are people who would prefer a more in town residence at old age due to convenience of reaching hospitals as well as not having to drive to reach anywhere meaningful, but that really depends on what sort of circle of life/friends one have built up before they age), but if answer to both questions are no, then it indicate it aren’t population at age driving the increase in population, which also mean those are people of the age of study/work

          • Herbert

            But the people of the age to study or work aren’t exactly getting more in Japan…

        • fjod

          You are, even by your own admission, proposing running duplex trains on lines with infrastructure that cannot support duplex trains.

          • fjod

            ‘support’ was michaelrjames’s wording but I assume this refers to the narrow loading gauge and resulting low bridges and tunnels. Quite expensive and time-consuming to upgrade these to Berne gauge for double-decker trains obviously.

          • Herbert

            It’d probably be easier to build a new line to a wider loading gauge and at that point, why not just relieve capacity constraints by building a new line, period?

          • fjod

            Can’t operate new lines with his condition that ‘they don’t involve more trains or more staff, ie. more operational cost’.

          • Herbert

            New lines also generate more ridership and hence revenue / social benefit…

          • fjod

            I was replying to michaelrjames’s comment. I don’t dispute that new lines can be beneficial.

        • Sassy

          It’s not really a profit motive thing. It’s more of a tradeoff between crowdedness and inconvenience. It seems like Japan prefers crowded convenient trains, than uncrowded inconvenient trains, so opportunities are taken to provide convenience via limited stop services and through running, rather than reduce crowding.

          Profit doesn’t have that much to do with it, if you look at the most crowded lines in Tokyo. Among the top 10, Den-en-toshi is the only line in the top 10 that is owned by a legacy private rail company. The Tozai Line operated by Tokyo Metro, which is government owned though for profit. The Nippori-Toneri Liner is operated by Toei, which is government owned and explicitly the government operator less concerned about profit than Metro is. The rest of the top 10 is JR East lines, which were more crowded in the JNR days.

          • Phake Nick

            >prefers crowded convenient trains, than uncrowded inconvenient trains
            Why do you think so?
            First of all, the crowdedness factor represent all trains across same line segement, no matter it is local train or express trains, so express trains’ load factor are also reflected in the crowdedness data.
            Second, express train is also a load balancing measure, and if the express trains are uncrowded while local trains are crowded, they will try to add more stops onto the express train in order to better distribute the load.
            Third, even if you cancel all the express trains, people will still need to get from point A to point B, the train carrying them can make all the stops but that simply waste everyone’s time more, including the train company’s operational resource, as with express train, the vehicles can complete one return trip faster and thus need fewer trains and fewer driver to provide the same service level.

          • Herbert

            Mixing speeds tends to be bad for capacity. You can’t run express trains without mixing speeds.

            Even if you rely on stuff like timed overtakes, you’ll still need safety margins and buffers…

            The maximum use out of any given infrastructure is achieved by running as homogenous a service on it as you possibly can – same acceleration and deceleration, same top speed, same stopping pattern, same everything… Of course that presents obvious drawbacks so in the real world various compromises are made and virtually no piece of rail infrastructure is maximized for capacity to the detriment of all else (interlining reduces capacity, but no sane person would propose stopping all Munich S-Bahn trains before the Stammstrecke and forcing people to change into trains that *only* do the Stammstrecke)

      • Phake Nick

        I don’t think any double tracked metro line in the world are capable of running at 40 train per hour.
        When China initially built their metro network, as their economy weren’t as developed at the time, they tried to save money by building short platforms, and hope to compensate for that by introducing more trains per hour. But as shown in reality none of them are able to reach like 40 trains per hours.

        Also, the proposal with completely separated deck, which will functionally be similar to double stack container trains, should have almost near double the original capacity, instead of the current double decker trains you see around the world.

        • Herbert

          https://en.m.wikipedia.org/wiki/V%C3%A9hicule_Automatique_L%C3%A9ger you can certainly argue those aren’t “proper” metro trains what with running on rubber tires, but those claim to (and apparently do) reach headways below 90 seconds..

          As for double stack container transport, it does not double carrying capacity per linear meter of train in virtually all cases – you see with the exception of, like, two lines in India you cannot stack two containers on a flatcar (American railroads will have you believe you cannot double stack under wire, but that is not true as Indian, Chinese and other operations can attest to) and instead need specialized “well cars” which require more space… An alternative to double stacking are so called “high cube” containers which have a height in excess of normal containers but smaller than two containers stacked atop one another. Then there are also slightly less tall containers which can be stacked under lower clearances but all those options have of course the downside of calling into question the advantage of containers that led to their development in the first place, namely their standardized dimensions and global interchangeability…

          As for a new type of double decker train, let me repeat here the concern about emergency egress… You’ll need enough stairs to get people out in an emergency and those need to be wide enough to allow for speedy evacuation…

          • Phake Nick

            Those AGTs are really really short trains, which is probably why they can run with less headway. Like some of those PRT are of similar concept.
            As for double stack container, the determining factor are loading gauge, whether and how easily would that fit in two levels of containers. With passenger trains it is also subjected to loading gauge, and will depends on how you layout the cars to fit in two layers as well as wheels and everything.
            And again, even for regular trains people need emergency stair or slider for leaving the trains in emergency, same can also provided for upper deck

          • Herbert

            “High floor” doesn’t mean “too high to jump out of”.

            In evacuations of aircraft (which happens via slide) including evacuation drills “a few broken bones” is generally considered an acceptable outcome if the time limit was maintained and everyone got out otherwise safely…

            Trains are like, what?, at most a meter from the rail to the entry door? That’s something that can be jumped by healthy adults and cynical as it may be, but the usual assumptions about mobility impaired people in the case of evacuation seems to be “well, good luck there, buddy”…

          • バニートルーパー (@archie4oz)

            I don’t think emergency egress is significant factor as a drawback for double-deckers.

            The bigger drawback is the additional weight which affects acceleration and deceleration (which impacts headway capacity), and track wear. Higher center of gravity which affects corner speed (again impacting headway capacity). Next month, the E4s are finally being retired from the Joetsu Shinkansen (being replaced by E7s), thus allowing them to substantially bump up the line operating speed.

            Another consideration is accessibility. Depending on the regulatory strictness, it does impose design and ingress/egress problems for the train, that may not be worth it.

          • Jojo

            There is no difference for max speed between single and double-deck trains on the french network (also remember that the fastest train ever is a double-decker). Only axle load and braking performance have an impact on this.
            Also, heavier doesn’t mean slower accelerating, the highest performance trains on the RER are the MI2N/MI09 because they have a higher power-to-weight ratio, simply.

            Accessibility is an issue, indeed. It’s part of the reason why both RER-NG and MI20 have single-deck cars.

          • Herbert

            We were talking about a double decker train that’s boarded on two levels (kinda like they tried with the A380 with separate jet bridges docking to doors on both levels; to my knowledge they couldn’t get it to work in practice) and the potential to remove passageways between levels in such cases (because pax board the level they want to be on anyway instead of boarding at the bottom and then climbing stairs if they want to go to the upper level) and I kept pointing out the possible safety issues this might cause in worst case scenarios…

          • Eric2

            Worst case scenario for a standard metro is already pretty bad – fall a meter onto the tracks, get electrocuted by the third rail, and hit by an oncoming train…

          • Herbert

            Which is why you remotely shut off the third rail and stop all train service on that line (plus, there are no oncoming trains if you have two separate single track tunnels even tho that is pretty rare globally speaking)

        • Jojo

          As Herbert stated, 42 trains per hour already works on 2 metro lines, and 40 tph on other lines, driverless or not.
          But the minimal headway depends on the time each train takes to enter the station, stop, start and clear it. So it depends on the dwell time (ie number of doors and people to get on/off), acceleration and braking performance, and train length.
          With very short trains, headways under 65 seconds or less are possible (look the VAL in Lille), but with a train more than 200 m long as on the RER, it has not been done under 2 minutes afaik.

          • Herbert

            So there’s a trade-off between length of vehicle and headway? Or is that a fake issue?

          • Jojo

            @Herbert. An example : when Toulouse doubled the length of VAL trains on line A (from 26 to 52m), the minimum headway grew from about 65 to 75 seconds. Of course this 13% reduction in frequency doesn’t compensate the doubling of the individual capacity.

          • Alon Levy

            Sure, but 26 m -> 10 s is a lot – it implies 2.6 m/s average speed during the acceleration and deceleration phases, so maybe 5 m/s top speed, which is very low.

          • Herbert

            I mean you can, in theory, achieve the same capacity by either running two car trains every five minutes or four car trains every ten minutes…

            Nuremberg notably runs (usually at least) shorter trains on U2 and U3 but at higher frequency than the longer trains on U1 (all platforms could in theory handle U1 length trains and frequently do – the newest generation of trains is four cars permanently coupled instead of two like the three preceding generations)

          • Jojo

            Yes, those 10 seconds look like a lot for only 26 m more. Maybe there are other issues like block, or delays added in the signalling system with its modernization (old systems like this are very fast, but up to modern safety standards…).

            Anyway, Paris’ subway line 14 is supposed to keep its 85-desconds headways with longer trains (130 m instead of 90).

          • Phake Nick

            In Hong Kong, after the upcoming extension of East Rail Line, the train will shrink from 12-cars to 9-cars long, in order to adopt to the geometry of the currently constructing, extended section of the line into Hong Kong Island, which cannot fit 12-cars train due to “limited underground space for train platform construction” and also fire safety protocol for undersea section.
            The MTR company claim that, together with signal system upgrade, it can boost the frequency of the train line from 23 tph up to 29 tph.
            Yet, 29*9 is still less train car and less capacity available than 23*12. I am not sure how they plan to deal with the extra ridership attracted by the train line extension, when currently there are already stations where passengers cannot board the immediate train in peak hour. They claim they hope another new rail line opened recently could help distribute the passenger load but that line go in a different direction and doesn’t really seems to be efficient in achieving what they want.

          • Eric2

            Hong Kong is making a lot of questionable transit decisions lately – this; lack of through running between East and South Island lines; lack of through running between Island and South Island West lines; lack of through running between Tung Chung and Tsueng Kwan O line…

          • Herbert

            Isn’t Hong Kong also a high construction cost place?

            As for the ten seconds for 26 meters thing… Maybe the old headway was unsustainable in practice and the decision was taken to increase it with the introduction of longer trains…

          • Mikel

            @Herbert

            So there’s a trade-off between length of vehicle and headway? Or is that a fake issue?

            It’s not the most common issue but in specific circumstances –branched systems with flat junctions–, the time that trains take to clear a junction can be a bottleneck on headways.

            One real-world example I know of is FGC Vallès in Barcelona. It runs 32 TPH peak from its 5-track terminal at Pl. Catalunya, with a branch point at Gràcia in which 8 TPH go to Av. Tibidabo and 24 to Sant Cugat; in turn, 12 of those continue on the Terrassa branch and the other 12 on the Sabadell branch (cf. Metro-North’s 24 inbound, 13 outbound at the 43-track GCT). The trains are pretty short (80 m) and there have been proposals to lengthen the platforms at great expense; one problem is that a few seconds’ increase in the time needed to clear the junctions (at low speed due to noise concerns in a very tight curve) would apparently compromise the stability of the schedule. That’s why the curently preferred solution is to build a completely new tunnel and through-run it to the FGC Llobregat line– it’s more expensive, but it would increase capacity more than twofold because the second pair of tracks would allow the segregation of local and express services.

          • Sascha Claus

            Length of the platform ought to have no effect on frequency, as the next train can (slowly) enter the station while the last one is still with its tail on the platform, like moving block with separation of only 10m. The following train has to be slow enough, but if it’s stopping anyway, that shouldn’t be a problem.
            Flat junctions or “Try to not stop inside the tunnel!” and “Do not enter the tunnel until the next station platform is clear!” are worse bottlenecks.

        • Herbert

          Of course weight has an impact on acceleration. As per physics, force is proportional to mass times acceleration and work/energy is force times distance, so more mass means more force for equal acceleration…

          Of course for virtually all passenger rail applications the real limit on acceleration is passenger comfort and unless we’re talking rollercoaster type safety features, those constraints affect acceleration way before those other concerns do – otherwise we’d be launching S-Bahn trains out the station with linear induction motors like a “launched coaster” https://en.m.wikipedia.org/wiki/Launched_roller_coaster

    • Herbert

      You still need staircases as means of emergency egress, don’t you? And given that in an emergency tone is of the essence, you can’t reduce their number and size too much, lest you create bottlenecks in the case of evacuation…

      • Phake Nick

        For regular metro cars running underground, the only viable emergency egress in usual situation are only the front and the back. So maybe you need some sort of folded up staircase like those emergency sliders attached to airplanes, but that’s probably sufficient already.

      • Phake Nick

        Also, given that suspended monorail is out there and running in the real world, which the only way of emergency egress is to open the floor-window and climb down, I don’t think too much worries are needed as double decker trains will most likely be able to evacuate faster than that.

        • Herbert

          The safety standards of an existing system (which is perhaps “grandfathered in”) do not justify jeopardizing the safety standards of a system yet to be built.

          Furthermore, it is generally frowned upon to replace existing systems with ones that are measurably worse in a measurable quantity. Not that this hasn’t happens (waves awkwardly towards bustitution) but usually not to high profile well connected citizens of the imperial capital…

          • Sassy

            Even though the technology is pretty unpopular, suspended monorails continue to be built. Germany built a new one in the early 2000’s, and China started building their first suspended monorails in the 2010’s and continue today.

          • Herbert

            http://www.pflichtlektuere.com/15/05/2012/h-bahn-unfall-an-der-mensabruecke/ this (article in German, because nobody outside the local area cares much about a few light injuries) shows there’s been an accident on the Dortmund H-Bahn which required evacuation which was effected by firefighting equipment (those ladders they have in their fire trucks 🚒). I’m pretty sure given the reaction time measured in minutes, this was a scenario thought of in the planning and approval process and the local fire Department is/was prepared for such a scenario.

            Now how do you propose to do something like that in a tunnel?

          • Sassy

            > Now how do you propose to do something like that in a tunnel?

            Several minutes is enough time to walk out the front door. More realistically in most subway fires, it’s enough time for the train to get to the next station. Relying on front door evacuation has been successful around the world, including on systems with excellent safety records.

  5. SB

    Is the global chip shortage having an impact on new train set orders?
    If so, that could explain some of the cost increase.

    • Herbert

      Why would it?

      I don’t think chips represent an appreciable share of the cost of trains, but do correct me if you have evidence to the contrary…

  6. Herbert

    Actually, the largest city in the EU (by a small margin) is Berlin ahead of Madrid – if one goes by official municipal boundaries, that is…

    I know, they have their own problems, but at least they are an identifiable line on a map, unlike stuff like “metro area” boundaries which add Frankfurt (Oder) to Berlin or Bielefeld to “Ruhr” (whatever that might be)

    • Alon Levy

      Wait, who counts Bielefeld in the Ruhr? I don’t even think it’s the same Verkehrsverbund. Frankfurt an der Oder and Berlin are in the same Verkehrsverbund, but there’s a separate definition of the Berlin Umland within Brandenburg that only includes actual suburbs like Potsdam and Oranienburg and has, with the city, 5 rather than 6 million people.

      At any rate, French metro areas are very strongly identifiable – France does contiguous urban development, like American and Asian cities and unlike German-speaking or Dutch ones. Nearly everything in Ile-de-France is part of the Paris urban zone, and the only reason Paris didn’t annex most of the suburbs in the 1950s is that they were working-class so the city didn’t want them and the state didn’t want to force the issue for fear Paris would get a communist mayor.

      • Herbert

        And now the suburbs hate Hidalgo’s anti car policies…

        How times change…

        Also, I’m not French but to my knowledge the only “boundary of Paris” that isn’t the municipal boundary that average French people recognize is the periph, isn’t it?

        • Alon Levy

          The Périph is the municipal boundary, more or less. And yeah, the suburbanites hate her anti-car policy, but nobody especially cares about their opinions.

          • Herbert

            Because they don’t vote in Paris elections…

            I’ve maintained for a while now, that you can actually do city tolls in a “populist” way if you communicate clearly enough that “only out of towners will pay, anyway” which of course is more or less what the CSU tried to do with their “Ausländermaut“ before the EU pointed out just how illegal that stunt was…

          • Phake Nick

            In Hong Kong when the government tried to push congested area car access toll, it’s usually the residents nearby who made the complain the it’ll inconvenient them, as they live there and have to drive in and out of it unlike other drivers who can just reroute through other roads.

    • Sassy

      And the largest city in Japan if one goes by official municipal boundaries is Yokohama. Municipal boundaries are always easy to define as they are legal boundaries, not real world boundaries.

      • Herbert

        https://en.m.wikipedia.org/wiki/Special_wards_of_Tokyo Tokyo is sui generis and there is actually a pre-existing historic boundary which was abolished by fascists (gee, bit direct with the “hating cities” thing, aren’t they?) But at any rate, official boundaries have the big advantage that you don’t have to explain where the line is

        What’s the value of having population figures for the same place at the same time that differ by millions of people according to opaque criteria that are never clearly spelled out?

        What’s more, who gets to draw the line of where “urban area” or “metro area”, or whatever the term is, ends? What do you do if there is some kind of “official” line for that, too, but you disagree with that same as you disagree with the official municipal boundaries? After all, there is no reason for a government to be sensible about one kind of border when they are eminently silly about another kind of border?

        And if you try to use the same term across different countries, you’ll inevitably have terms that at the very least sound similar to something someone else already did… In short, chaos… And nobody wants to read thru pages of discussion on the ontological entities of cities before being given an answer as to which city is how big…

        • Sassy

          > Tokyo is sui generis

          Not really? For example, Seoul and Bangkok have municipal governments and boundaries at the ward/district level as well, and I’m sure there are more.

          > What’s the value of having population figures for the same place at the same time that differ by millions of people according to opaque criteria that are never clearly spelled out?

          What’s the value of having population figures for a priori drawn arbitrary polygons on a map?

          And the criteria need not be opaque or vague. There are studies which define criteria and go out measuring based on them. Of course if you’re relying on commonly accepted or official criteria for metropolitan areas they vary from region to region, but this problem affects municipal boundaries much heavier than it affects commonly accepted/official metropolitan area boundaries.

          At least with official/commonly accepted metropolitan area boundaries, the goal of what is to be bounded is more clear and consistent than municipal boundaries, so less nonsense results occur.

          Paris is without a doubt the biggest city in the EU, and Seoul is without a doubt the biggest city in Korea. Going by municipal boundaries would contradict the truth, so municipal boundaries are not appropriate for the question of “how many people live in X city”

          • Herbert

            Who does the common accepting and on what basis?

            I don’t remember being asked whether Hof an der Saale is “A suburb of Nuremberg” or not. Were you asked?

          • Herbert

            Like, if there are easily applicable criteria, what are they? What are they based on? Who decides on them? Who authorizes those who decide on them?

          • Herbert

            Tokyo is sui generis within Japan, just like D.C. is sui generis within the U.S., D.F. is sui generis within Mexico and so on…

          • Sassy

            > Who does the common accepting and on what basis?

            Some collective consciousness I guess? Despite your insistence that the measure is vague, almost every time someone talks about the population of the Paris metro area they use Île-de-France boundaries, and almost every time someone talks about the population they use Itto Sanken boundaries. There are definitely other boundaries one could use, but it’s hard to argue that there isn’t any commonly accepted boundary.

            > Like, if there are easily applicable criteria, what are they? What are they based on? Who decides on them? Who authorizes those who decide on them?

            Worldwide consistent criteria is generally just decided by authors of a study, and then the study is published, and people can peruse through many metropolitan areas bounded by the same criteria. e.g., The Atlas of Urban Expansion.

            > Tokyo is sui generis within Japan, just like D.C. is sui generis within the U.S., D.F. is sui generis within Mexico and so on…

            Ah look at another example proving the non-uniqueness of Tokyo (though D.C. isn’t really one since I don’t think there are several municipal governments within D.C.).

            Why does it matter that Tokyo is unique within Japan? Other cities are also eligible to switch to municipal level wards, and it even came up for vote for Osaka in 2015 even if it didn’t pass.

            If you’d count D.C. in the list, then Bangkok is not unique within Thailand either, as Pattaya is similar to D.C.. And also within Thailand, there has been talks to restructure Chiang Mai like Bangkok.

          • Sassy

            Oh, and Seoul is far from unique in South Korea (I think?), as plenty of other cities like Incheon and Busan are province level, with their districts at municipal level.

          • Herbert

            Isn’t Ille de France just another administrative boundary? I thought those were double-plus-ungood?

            Also: Berlin is state level and has elected sub-governments (“Bezirke”) which only don’t count as “cities” because they are not called that…

          • Alon Levy

            Ile-de-France is a region, with the same powers as any other region. The main differences are,

            1. Ile-de-France has money, so things that are more devolved, like education, have more money and then everyone brags about the Sorbonne (non-selective unis in France – including every uni that has “Sorbonne” in its name – admit by region.)
            2. Some things in Ile-de-France have national significance, so the state takes more interest, which plays a role in the region’s recent housing growth.
            3. The national elites live in and around Paris, so national outfits like SNCF are more responsive to the needs of Paris than to those of secondary cities.
            4. Regional elites elsewhere aim to imitate Paris, minus the part where Ile-de-France is 30% people of color (Marseille has anti-kebab zoning).

          • Sassy

            > Isn’t Ille de France just another administrative boundary? I thought those were double-plus-ungood?

            As I mentioned previously, official criteria for metropolitan areas they vary from region to region, but this problem affects municipal boundaries much heavier than it affects commonly accepted/official metropolitan area boundaries, as at least with official metropolitan area boundaries, the goal of what is to be bounded is more clear and consistent than municipal boundaries, so less nonsense results occur.

            > Also: Berlin is state level and has elected sub-governments (“Bezirke”) which only don’t count as “cities” because they are not called that…

            I’m not familiar with Berlin, but if the city is the state and the Bezirke are municipalities, then going by municipal borders, Berlin wouldn’t be the largest city in Europe either.

          • Herbert

            The Bezirke of Berlin aren’t municipalities, they just behave and function as if they were – including boundaries drawn in part out of administrative considerations first and foremost…

            But they are not municipalities. The state of Berlin is the city of Berlin and that’s the end of that.

  7. Herbert

    Why wouldn’t the doors on the sides provide emergency egress? You’ll need space enough besides the train for people to move thru the tunnel anyway for the smaller kind of maintenance and inspection done during normal operation. Plus, I don’t think metro trains usually come with a door that opens in the direction of travel – and if they do, passengers are not aware of them…

    Also, no metro tunnel I know of has the height clearance for sensible double decker designs – certainly not if you add the extra space needed for a platform (which needs a certain thickness to be able to bear the load of all the waiting pax)

    Anyway, we’re largely talking about S-Bahn style operations here which spend the vast majority of their route km above ground…

    • Herbert

      This comment has been threaded wrong, it should be a response to “phake Nick”

    • Sassy

      > I don’t think metro trains usually come with a door that opens in the direction of travel – and if they do, passengers are not aware of them

      Evacuation doors in the direction of travel are the norm for any train that goes into a subway in Japan, and they show up in Korea, Taiwan, China, etc. as well.

      Whether people notice them, idk. Probably not, but in an orderly evacuation, they would be instructed to go through that door.

      • Herbert

        If your evacuation plan assumes “orderly” or even just “personnel will be able to give instructions” that’s already an ill conceived evacuation plan that’s blind to Murphy’s law…

        Basically, if it takes you less than five minutes to come up with ways this could go wrong, you’re not doing enough…

        • Sassy

          Not all evacuations are rushing to get out of the train. I’d go as far as to say the vast majority of evacuations aren’t.

          Front evacuation doors work fine in a lot of the world, and are used in systems with particularly high concern for safety and good safety track records.

          • Herbert

            When you need an evacuation under the worst possible conditions, the fact that evacuations under better conditions work just fine and are more common anyway won’t help you…

            Like, not everything is Kaprun, but can we prevent the next one before a hundred people die, please?

          • Sassy

            Why do the worst possible conditions have to be handled by better evacuation? It’s often much better ROI to handle them by making them much more rare.

            For example, instead of building trains with higher buff strength, work towards eliminating situations where that might be useful.

            People don’t design airliners such that every seat ejects out in an emergency. Many lives could be saved if they did. I wonder why…

          • Herbert

            Do you know the survival rates of emergency ejections in fighter jets? And those are trained pilots. As the old adage in seafaring went “Often the best rescue craft is the original vessel”… And in most cases where an ejection seat might be a consideration (and good luck having some schmuck getting ejected while not expecting it) there are other things that have a higher chance of success – such as gliding or landing on whichever roughly clear patch of dirt or water that can be reached…

            And in case you were unaware: The number of so called “zero to zero” ejection seats is very low – and they subject their occupants to very heavy accelerations, which, again, are not a picnic even for trained fighter jet pilots…

            Meanwhile “train gets stuck in a tunnel, burns” is not an “out there” scenario and the obvious solution “Allow people to get out thru the doors and walk to safety thru the tunnel” should at least be considered…

          • Sassy

            > Meanwhile “train gets stuck in a tunnel, burns” is not an “out there” scenario and the obvious solution “Allow people to get out thru the doors and walk to safety thru the tunnel” should at least be considered…

            Why? Even if they got out of the sides of a burning train, is it even possible to walk out of the tunnel alive, considering they’d have to be walking right next to a burning train? In a tunnel where the fire might consume most of the oxygen? And if there’s plenty of space around the train inside the tunnel to mitigate those issues, then shouldn’t you just save some money and use all that extra space to make the tunnel double track instead of single track?

            Front evacuation doors to mitigate the risk of having to evacuate people from a train stuck in a narrow tunnel are used in many of the safest systems in the world. And even when accidents happen in those systems, they would have not been mitigated by ensuring all tunnels have enough space for side door evacuation.

            If anything, less safe systems that ensure that all tunnels are wide enough for side door evacuation, are just wasting resources, since as the safer systems show, there is much lower hanging fruit to be picked wrt safety.

            > Often the best rescue craft is the original vessel

            The newer Shinkansen trains have battery backups to get people out of tunnels in the original vehicle, and in some cases even enough to limp back to a station. With improvements in battery technology, that might be rolled out to commuter trains in the future as well.

            As often the best rescue craft is the original vessel, this approach could save many more lives than making tunnels wider.

          • Herbert

            The longest tunnels in the world all have some (often pretty elaborate) systems to force fresh air into places and to remove smoke to allow breathable air. Of course for road tunnels this is utterly non-negotiable because motorcars already produce choking hazardous fumes under “normal” operating conditions… Such systems usually come with double and triple redundancy for obvious reasons.

            There are also shelters (with their own air supply and a slight overpressure keeping out outside air) as well as means to get to the other tube in twin tube tunnels.

            Those are all pretty standard features in all new builds in Europe and have been for a while now. Why do you insist they are all a giant waste and safety doesn’t matter?

            Also, good luck getting a train out the Gotthard Base Tunnel if the fault that caused the fire is in the main propulsion system…

          • Sassy

            > Those are all pretty standard features in all new builds in Europe and have been for a while now. Why do you insist they are all a giant waste and safety doesn’t matter?

            It’s not that safety doesn’t matter. Safety does matter, and based on what we see in the real world, more lives would be saved if money was spent elsewhere. A train catching fire in a single track tunnel is an extreme edge case that should be handled after much bigger killers such as level crossings, and people falling/jumping off of platforms. If for whatever reason trains regularly catch fire in single track tunnels, then the solution is to stop that from happening, because it clearly can be done, since it has been done.

            If you add in the health risks of cars, transit safety can be argued is taken too seriously, and money spent displacing cars will save far more lives than money spent making current transit users safer. However, the most effective transit safety measures, such as good maintenance, fewer level crossings, better signalling, automation, and platform screen doors, also make transit and cities overall better and are worth doing, even if one is against safety theater.

          • Herbert

            https://en.m.wikipedia.org/wiki/Channel_Tunnel_fire if fires in tunnels are an extreme edge case nobody needs to worry about, why is this a disambiguation page linking to several incidents in one high profile tunnel alone?

            Also, as per my previous comments, the design of the Channel Tunnel (two separate single track tubes as well as a “service tunnel” between them linked all 350 meters which has been basically copied by all major rail tunnels since for a reason) works. What could’ve easily have resulted in a
            https://en.m.wikipedia.org/wiki/Mont_Blanc_tunnel_fire Montblanc tunnel or Kaprun sized https://en.m.wikipedia.org/wiki/Kaprun_disaster conflagration consuming human bodies like so much sacrifices to a perverse idol instead ended up being a mostly annoying case of property damage easily repaired in the ensuing days and weeks…

          • Sassy

            > https://en.m.wikipedia.org/wiki/Channel_Tunnel_fire if fires in tunnels are an extreme edge case nobody needs to worry about, why is this a disambiguation page linking to several incidents in one high profile tunnel alone?

            Maybe the tunnel should be operated better then? If fires in narrow single track tunnels were common and front door evacuation is insufficient, you’d expect people to have died because of the design “flaw” in those systems, but they haven’t.

            > Also, as per my previous comments, the design of the Channel Tunnel (two separate single track tubes as well as a “service tunnel” between them linked all 350 meters which has been basically copied by all major rail tunnels since for a reason) works.

            The Chuo Shinkansen tunnels use the typical design with a single double track main tunnel. If it was so important to copy the Channel Tunnel design, you’d think a company interested in protecting their perfect safety record operating in one of the most natural disaster prone places in the world would adopt it, but they didn’t.

            > Montblanc tunnel

            A car tunnel

            > Kaprun

            How does that prove the necessity of side evacuation doors? Even though there were problems along the way, the side doors did open, and many people did evacuate through them, to their deaths. The only people that survived escaped by breaking the window at the end of the train. More people would have survived if there was an evacuation door there.

            The official investigation even stated “rescue or self-rescue concepts, as provided in newer railway or road tunnels [i.e., the stuff of which are advocating for], do not form an appropriate basis for the system design of tunnel funicular railways” and focused on better fire monitoring and fire extinguishing. There was outrage at the report of course, but I’m pretty sure it was at the people who chose to run dangerously modified equipment with inadequate maintenance going unpunished.

          • michaelrjames

            @Sassy

            The big killer in fires is always the smoke not the heat and fire itself. The best way to escape the smoke–if moving the train to the next station or out of the tunnel cannot be done–is to exit the train from the opposite end to the fire and walk up the tracks. Exiting the sides slows down this escape and often puts people right into the smoke zone (as it seeks a way past the train) even if there is a narrow walkway there. The train itself slows down the spread of smoke and so exits at its ends makes most sense.

          • Tom the first and best

            The Kaprun fire was in the lower drivers cabin, at the end of the train. Evacuating through the upper end would only have increased the risk of passengers evacuating upward, which is what killed the passengers who got out of the train through heat/smoke.

            The survivors were in the compartment next to the fire so they started trying to get out (of the side of the train) faster, succeeding faster and they included someone with 20 years experience as a volunteer firefighter who made them evacuate downwards past the fire to safety bellow the heat and smoke.

          • Sidney

            Its weird how people can say something as a fact then assert the exact opposite…

            If “Often the best rescue craft is the original vessel”, why applying this principle to trains is unsafe?

      • Jojo

        All trains running in Paris can only be evacuated from the side. It has happened from time to time on the metro.
        Usually the tunnels have some sort of narrow sidewalk on one side of the train so people can walk there.

        • Herbert

          Are those “sidewalks” also used for routine inspection / maintenance while trains are running?

    • Phake Nick

      >Why wouldn’t the doors on the sides provide emergency egress?
      Because metro run mostly underground inside tunnel, and there usually aren’t enough space between the door and the tunnel’s wall to let passengers step down, stand up, and then walk

      • Herbert

        That depends on the design of the tunnel. Most metros aren’t the Tube – heck, even much of London underground has more space between doors and walls than the iconic “Tube” images…

        • Phake Nick

          Even a Shinkansen-sized tunnel cannot allow such thing happen. Which is why they are now adding batteries onto new Shinkansen trains, providing power for the train to drive out of tunnel in case of power outage after earthquake, so as to prevent people from trapped inside trains inside tunnel.

          • Herbert

            What are you talking about? All the newer European high speed rail tunnels are built as two separate tubes with frequent links between the two so that in case of emergency one can serve as the “emergency exit” of the other…

          • Herbert

            Thru the side doors. You can’t possibly evacuate a 400 meter train thru a “front door” even if one existed…

            There have been evacuation drills (as well as a handful real life evacuations) of various trains in various tunnels – pictures of that can be found online

          • Herbert

            Also, tunnel boom and the piston effect make it not advisable to make tunnels too narrow anyway…

          • Sassy

            Most Shinkansen tunnels are a single hole with two tracks. While I wouldn’t be too surprised if narrow single track tunnels exist somewhere in the system, in nearly every emergency stop situation, if there was a real need, people could exit the train on to the other track.

            The advantage of the battery backup is that the train itself can leave the tunnel. Being out of the train but still in the tunnel is useless in many disaster scenarios, so being able to get the passengers out of the tunnel in the train in an emergency could be useful.

          • Herbert

            The Seikan tunnel even has two stations whose main purpose seems to be emergency evacuation – and apparently they’re not big enough to allow all door boarding, so at least part of their use case must be pax arriving per pedes…

          • Sassy

            The Seikan Tunnel is a double track tunnel, which as previously mentioned, inherently allows people to leave the train from all doors if the need arises.

    • Eric2

      Regional rains have relatively long rides, so when it gets crowded you have to provide more seats to maintain a tolerable comfort level, which means double deckers.

      Metro trains have relatively short rides, so when it gets crowded standing is more acceptable, so double deckers are not used.

      • Herbert

        And if you find yourself having a long ride on a train that makes a lot of stops, you are either some kind of penny pincher (in cases where the fare is different for different kinds of train) or the system has an inadequate amount of “express” options…

        • Jojo

          Well there are just no tracks to run an “express” service along the southern half of RER B. And not much space do create them unless we dig a completely new line.

          • Herbert

            Is there no parallel mainline or Transilien line thereabouts?

          • Alon Levy

            No, the RER B’s southern portion descends from a commuter line that was never really part of the national intercity network and went to a dedicated terminal at Luxembourg, which is how it got incorporated into RATP early. It does run some express service, but within the city all trains make all stops, and the express trains aren’t especially fast even in the suburbs.

          • michaelrjames

            There is nothing that running these new duplex trains won’t fix about the RER-B line. The stations are already adequately spaced, especially in its inner portions (Petite Couronne) given that it is fairly densely urbanised. That is, an express doesn’t make any sense, quite apart from being politically out of the question. In fact, the richest parts, Sceaux, effectively have an express service by virtue of their two branches and 5 stations! It is a very old suburb, a royal one based around the chateau and its parc, and as Alon indicated, was one of the earliest to have suburban rail into the centre, ie. to Luxembourg. This is where the Curies lived with Marie commuting by rail to Luxembourg which is a short walk to her institute (today’s Curie Institute, 2 blocks south of the Pantheon).

            @fjod
            En passant, earlier I believe I said that running duplexes on the Southern lines in England would have solved their problems. But of course they can’t run such trains and I wasn’t particularly suggesting they should re-engineer the lines, though I don’t know how much work would be required. Apparently the additional vertical clearances involved are not so great; I’ve never understood why the base of existing tunnels can’t be excavated the relatively modest amounts to provide the clearances. And in piecemeal overnight works while keeping the line operating as usual. Oh, and I’d guess it could have been funded just from all the additional state subsidy given to the private franchise operation or maybe just from the bonuses awarded to themselves by upper management. I’m sure any official reports into this issue by the usual suspects, in the last 50-70 years (when the southern lines feeding London became very busy) have come to the conclusion it was too expensive and disruptive etc but then the entire political system has been either hostile to rail and/or hostile to spending money on public transit. I’m currently reading Nicholas Faith’s “The History of the Channel Tunnel” in which he explains that it only got built because Thatcher had been convinced it could be a drive-thru tunnel, and indeed this absurd distraction (every engineer knew it didn’t make any sense) was kept alive throughout the design & tendering process, right up to decision making, for this exact reason. To be fair and not simply anti-Thatcher or anti-British-establishment, Faith claims that Mitterrand was of similar mind and that helped Thatcher’s support of the Tunnel (with proviso of zero public money) I can’t help be a tad sceptical of that claim, especially as he would have received more direct advice from the more politically-influential French engineering fraternity.

          • Herbert

            There’s always a use case for a more express service.

            The question is: how big is that use case and what kind of infrastructure does it justify?

            DB built their Montabaur and Limburg stops on the Cologne Frankfurt line to be easily bypassed at 300 km/h for a reason and trains that don’t stop there are at least as frequent as those that do…

            But otherwise the ICE Sprinter is basically a rounding error, no matter how much people complain about trains stopping at “milk jugs” (“Milchkannen”)

            That may well be due to the overtaxed German rail system which would require expensive bypasses to allow Sprinters to actually be notably faster, but then even when there seems to be an obvious “bypass” option (e.g. Avoiding the Leipzig dead-end by serving LEJ airport or Halle instead) it is rarely taken…

            Of course France seems to have a different approach – there, it would appear, every Frenchman demands the right to get to Paris non-stop from the provinces, but the unfortunate schmock who has to go from Provincial town a to Provincial town b can try getting a cab from one of Paris’s seven bazillion dead end stations to the other – if the trains even remotely align temporally for anything approaching a connection…

          • Herbert

            I don’t know the rail lines in southeastern England, but I do know (both from having been a passenger and from reading about it) the Right Pegnitz Valley line. (“Right” as in “orographically to the right of the river in direction of flow” – to distinguish it from the left Pegnitz Valley line, part of which is electrified and used by S1).

            You see, they want to electrify that line (it is both the shortest connection by rail from Nuremberg to Cheb and part of the important links from Hof (onwards to Leipzig/Chemnitz/Dresden – all already electrified) to Nuremberg, Regensburg and points south and east. It would thus form both an important pax and freight connection in its own right (it was one of the lines once served by “Diesel ICE”) as well as an important backup for already overtaxed north-south links further west…

            Anyway, they want to electrify that line per https://www.bahnausbau-nordostbayern.de both for S-Bahn and for long distance service, but as the region is, well… Hilly is probably the word… And the Pegnitz as a pretty winding river represents pretty much the only even remotely flat route thru there are more bridges and tunnels than one can count… Now, I don’t care that some of those bridges are heritage listed (apparently neither does DB, but they try their darndest pretending) but that doesn’t change geography or physics… Oftentimes, the sequence is (without really any “open track” in between) “bridge, tunnel, bridge, bridge, tunnel” so if you try to get the necessary height clearance (the tunnels, of course, were built to the cheapest possible loading gauge) by lowering the floor of the tunnel (which requires shutting down traffic to move the rails out of the way, dig and install rails again, but anyway) you’ll have to move the bridges the same amount lower, too – and many of those bridges simply can’t be lowered. It’s not unsolvable, smart DB people have come up with solutions on the abovementioned website, but it is much harder than it would at first appear.

            As an aside, I wrote some comments on this thread while riding on that rail line on Sunday…

          • michaelrjames

            Herbert: ” by lowering the floor of the tunnel (which requires shutting down traffic to move the rails out of the way, dig and install rails again, but anyway) you’ll have to move the bridges the same amount lower, too – and many of those bridges simply can’t be lowered”

            It would be nice to hear what an engineer thought. But my main point, which eludes you, is that mostly the difference apparently is not much, like 10-20cm. So the bridge-tunnel-bridge sequence could surely handle a bit of this gentle ‘roller coaster’. Isn’t that partly how the original European HSR line, Paris-Lyon, manage to be so cheap to construct? As to shutting down traffic etc, we had this same discussion concerning installation of Paris T3 tramway, which was installed across its approx. 30 major street arteries by using prefab units that could be put in place overnight and take road traffic hours later. That’s the way its 35km was built and apparently never requiring daytime interruption of traffic on those arterials (leading across the Peripherique). If we’re talking mere 10-30cm, the tunnel’s base can be excavated, new (quick-set) concrete poured (or pre-made bits dropped into place) overnight and the original rails reinstalled for BAU next day. Not until the whole tunnel is done is the whole lot reinstalled with new lower track. It really doesn’t strike me as rocket science.

            In the UK it is almost certainly a lack of willingness (often on the pretext of ‘no money’). Why is it that the UK stands out amongst rail nations in being so crippled by its unchanged legacy lines? It seems they refuse to change anything unless in extremis they build some completely new zillion-dollar thing like HS1 or Crossrail? Why/how is it that most of France’s long-distant rail, and 5 of 5 RER lines in Paris can take those duplexes (even if B required custom trains; Alon can complain about cost but the real cost of not doing such things is very visible in those UK Southern commuter lines)? Probably simply because they wanted them to. Oh, and it is simpler to do this than rebuild lines to take wider trains or longer trains (which it did of course with twin TGVs but which wasn’t enough increase in capacity). It would be informative to know (JoJo, Onux, fjod, any engineers) what remedial works the French did. Heck, they even run duplex RERs thru the ancient system of the PC–right bank, partly the former Auteil-St-Lazare commuter line from 1854!.

          • Alon Levy

            UK rail electrification costs are pretty reasonable when you subtract out bridge modifications. The UK loading gauge is uniquely restricted, so bridge mods that aren’t usually needed here are needed there.

            The real problem is when American railroaders think they have restricted clearances too when their loading gauge is way higher than what we have here and then British engineers believe them instead of calling bullshit.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.