FRA Reform is Here!

Six and a half years ago, the Federal Railroad Administration announced that it was going to revise its passenger train regulations. The old regulations required trains to be unusually heavy, wrecking the performance of nearly every piece of passenger rolling stock running in the United States. Even Canada was affected, as Transport Canada’s regulations mirrored those south of the border. The revision process came about for two reasons: first, the attempt to apply the old rules to the Acela trains created trains widely acknowledged to be lemons and hangar queens (only 16 out of 20 can operate at any given time; on the TGV the maximum uptime is 98%), and second, Caltrain commissioned studies that got it an FRA waiver, which showed that FRA regulations had practically no justification in terms of safety.

The new rules were supposed to be out in 2015, then 2016, then 2017. Then they got stuck in presidential administration turnover, in which, according to multiple second-hand sources, the incoming Republican administration did not know what to do with a new set of regulations that was judged to have negative cost to the industry as it would allow more and lower-cost equipment to run on US tracks. After this limbo, the new rules have finally been published.

What’s in the new regulations?

The document spells out the main point on pp. 13-20. The new rules are similar to the relevant Euronorm. There are still small changes to the seats, glazing, and emergency lighting, but not to the structure of the equipment. This means that unmodified European products will remain illegal on American tracks, unlike the situation in Canada, where the O-Train runs unmodified German trains using strict time separation from freight. However, trains manufactured for the needs of the American market using the same construction techniques already employed at the factories in France, Germany, Switzerland, and Sweden should not be a problem.

In contrast, the new rules are ignoring Japan. The FRA’s excuse is that high-speed trains in Japan run on completely dedicated tracks, without sharing them with slower trains. This is not completely true – the Mini-Shinkansen trains are built to the same standards as the Shinkansen, just slightly narrower to comply with the narrower clearances on the legacy lines, and then run through to legacy lines at lower speed. Moreover, the mainline legacy network in Japan is extremely safe, more so than the Western European mainline network.

On pp. 33-35, the document describes a commenter who most likely has read either my writings on FRA regulations or those of other people who made the same points in 2011-2, who asked for rules making it possible to import off-the-shelf equipment. The FRA response – that there is no true off-the-shelf equipment because trains are always made for a specific buyer – worries me. The response is strictly speaking true: with a handful of exceptions for piggybacks, including the O-Train, orders are always tailored to the buyer. However, in reality, this tailoring involves changes within certain parameters, such as train width, that differ greatly within Europe. Changes to parts that are uniform within Europe, such as the roofing, may lead to unforeseen complications. I don’t think the cost will be significant, but I can’t rule it out either, and I think the FRA should have been warier about this possibility.

The final worry is that the FRA states the cost of a high-speed train is $50 million, in the context of modification costs; these are stated to be $300,000 for a $50 million European high-speed trainset and $4.7 million for a Japanese one. The problem: European high-speed trainsets do not cost $50 million. They cost about $40 million. Japanese sets cost around $50 million, but that’s for a 16-car 400-meter trainsets, whereas European high-speed trainsets are almost always about 200 meters long, no matter how many cars they’re divided into. If the FRA is baking in cost premiums due to protectionism or bespoke orders, this is going to swamp the benefits of Euronorm-like regulations.

But cost concerns aside, the changes in the buff strength rules are an unmitigated good. The old rules require trainsets to resist 360-945 metric tons of force without deformation (360 for trains going up to 200 km/h, 945 beyond 200 km/h), which raises their mass by several tons per cars – and lightweight frames require even more extra mass. The new ones are based on crumple zones using a system called crash energy management (CEM), in which the train is allowed to deform as long as the deformation does not compromise the driver’s cab or the passenger-occupied interior, and this should not require extra train mass.

How does it affect procurement?

So far, the new rules, though telegraphed years in advance, have not affected procurement. With the exception of Caltrain, commuter railroads all over the country have kept ordering rolling stock compliant with the old rules. Even reformers have not paid much attention. In correspondence with Boston-area North-South Rail Link advocates I’ve had to keep insisting that schedules for an electrified MBTA must be done with modern single-level EMUs in mind rather than with Metro-North’s existing fleet, which weighs about 65 metric tons per car, more than 50% more than a FLIRT per unit of train length.

It’s too late for the LIRR to redo the M9, demanding it be as lightweight as it can be. However, New Jersey Transit’s MultiLevel III is still in the early stages, and the railroad should scrap everything and require alternate compliance in order to keep train mass (and procurement cost) under control.

Moreover, the MBTA needs new trains. If electrification happens, it will be because the existing fleet is so unreliable that it becomes attractive to buy a few EMUs to cover the Providence Line so that at least the worst-performing diesels can be retired. Under no circumstance should these trains be anything like Metro-North’s behemoths. The trains must be high-performance and as close as possible to unmodified 160 km/h single-level regional rail rolling stock, such as the DBAG Class 423, the Coradia Continental, the Talent II, or, yes, the FLIRT.

Metra is already finding itself in a bind. It enjoys its antediluvian gallery cars, splitting the difference between one and two decks in a way that combines the worst of both worlds; first-world manufacturers have moved on, and now Metra reportedly has difficulty finding anyone that will make new gallery cars. Instead, it too should aim at buying lightly modified European trains. These should be single-level and not bilevel, because bilevels take longer to unload, and Chicago’s CBD-dominant system is such that nearly all passengers would get off at one station, Millennium Station at the eastern edge of the Loop, where there are seven terminating tracks and (I believe) four approach tracks.

Ultimately, on electrified lines, the new rules permit trains that are around two thirds as heavy as the existing EMUs and have about the same power output. Substantial improvements in train speed are possible just from getting new equipment, even without taking into account procurement costs, maintenance costs, and electricity consumption. Despite its flaws, the new FRA regulation is positive for the industry and it’s imperative that passenger railroads adapt and buy better rolling stock.


  1. alexhutcheson

    I had to double-check to make sure this wasn’t some variation of your annual April Fools posts.

    • Alon Levy

      Check the date.

      (Unfortunately, my 2016 post was so close to reality that some people thought it was real; I had to chase down media sources explaining to them it was a parody. That some of it has since become real has not helped.)

  2. adirondacker12800

    making it possible to import off-the-shelf equipment.

    There isn’t any. Raise the floor another few hundred millimeters, widen the body a few hundred millimeters then lengthen the whole shebang a few meters, you have to redesign the whole thing. Once the MTA, NJTransit or Metra bites the bullet, SEPTA, MARC, the MBTA, AMT, Go and any other obscure tiny little operators can say “We’ll take some of those, with different paint and upholstery” .

    • Alon Levy

      They already raise the floor for British orders (and also for some older high-floor equipment with steps from platform level) and widen the carbody for Scandinavian ones. Swedish train dimensions are similar to American ones.

      • adirondacker12800

        And how many M7/M7a’s are there? Compared to all the passenger cars in Sweden? The MTA or NJTransit announces they want something there’s going to lots of competition. When CDOT electrifies to Springfield they are gonna order up some more M10s not go searching for something new and exciting. Or CDOT and the MTA will use the Springfield line to figure out an M12. That’s going to be sometime around the time Denver is thinking about new cars for their clapped out Silverliners. They’ll have Silverliner VIIs or M10s or Mutlilevel IIIs to pick from. Sweden won’t be particularly relevant.

    • Max Wyss

      A good design allows for such modifications without having to recreate everything. That’s the advantage of the various product families (not only from Stadler, but Bombardier and Alstom have similar products).

      • adirondacker12800

        Like the design Rotem used for SEPTA in Philadelphia that they sold to Denver’s RTD? Or Bombardier’s designs that got modified for NJTransit and then got sold to Montreal’s AMT and Maryland’s MARC, with … different paint and upholstery? You do realize that foamers get all frothy this time of year when they can go down to the station and watch MARC cars go through the SEPTA station. Or NJTransit cars. The videos should be on YouTube by Friday. Maybe even earlier, I don’t want to bother to look up if they are running “extras” on Amtrak’s schedule today. When NJTransit or the MTA decides to make a move, the usual suspects aren’t going to attempt to import stuff. Or use a design that needs major modifications to work with North America’s loading gauge. They are going to use their existing North American assembly plants and supply chain. They sell something to a Northeastern operator, almost anyone else in North America…needs to change the paint and upholstery.

        • F-Line to Dudley

          The “Brokem” Silverliners are a completely custom design effed up to hell by a micromanaging SEPTA and an incompetently managed factory. Denver RTD held its nose and tagged along with an option order of rote SEPTA stock (sans 25 Hz transformer core) because they got a steal on unit price and delivery dates best matched their system opening. It’s a dead-end lineage because all the chronic problems and quality control issues those vehicles and the Rotem coaches for the MBTA, Metrolink, and TriRail have had pretty much have gotten Rotem blackballed from any more domestic orders for the time being. When it’s time to replace the Silverliner IV’s SEPTA is going back to the design drawing board all over again, and Denver’s next fleet expansion will be something different.

          Bombardier’s BLV (8-inch boarding bi-level) is the closest thing to an off-shelf vehicle there is. Most widely used coach on the continent, in-service since 1976 with 8th generation update now in-production, excellent reliability and rebuildability rep, and at 110,000 lbs. a little bit lighter than an MLV (132-139K). About the only knock against them is that Rotem bungled its chance to open up a market for direct clones with its poor-quality Metrolink and TriRail orders, so at this moment BLV’s are still single-manufacturer sourced from a company that’s in lots of internal turmoil. Pullman gallery cars are probably the next-most off-shelf design, as they’ve been in-production with a whole slew of clone vendors for over 60 years. Light mutations on the design with each new manufacturer that’s dabbled in them, but similar enough overall that lots of agencies which prefer the gallery setup have started out with ancient refurbbed Metra beaters then graduated seamlessly to new orders.

          That’s about it, however. Formerly near- off-shelf Comet II-IV aluminum single-levels haven’t been produced in over 20 years, with the Alstom Comet V’s–the last commuter rail flats produced in North America–being a botched stainless steel redesign that’s overweight with notoriously rough ride quality. Dead-end lineage. MBTA/MARC Kawasaki bi-levels haven’t been produced in 10 years, with Rotem bungling the attempt at a clone market and Kawasaki uninterested in producing any new batches. And it remains to be seen what happens with the MLV, as SEPTA and AMT/Exo have opted for all-new design bi-level orders from CRRC instead of buying the NJT-spec MLV’s. NJT will keep ordering them, but unless the MTA follows suit ordering MLV’s with its combo MNRR/LIRR diesel coach procurement the market for 48-inch boarding coaches is going to be pretty thoroughly fragmented unlike their 8-inch counterparts.


          The MultiLevel EMU is somewhat intriguing because Bombardier’s intent in going for that NJT order is to be able to serve up the same propulsion package in different standardized frames. If they can get a reliable-enough one in MLV packaging, they’ll immediately offer up the same vehicle for GO Transit’s electrification in 8-inch boarding BLV packaging. And in each case the EMU version is supposed to be able to trainline with a stock MLV or BLV dead trailer that’s re-wired for MU circuitry, meaning existing coach fleets (numbering in the thousands) can be used. While inducing somewhat of a performance penalty, that does take the edge off the up-front cost of electrification vehicle procurements and could be a cost/risk assessment breakthrough for agencies thinking of taking the first-time EMU plunge.

          The downside is that trainling with alike coaches requires starting with a very heavy frame for the self-propelled version, so the EMU MLV’s are going to be very porky and probably limited in weight-shedding redesign potential because of the need to trainline with pre-existing stock coaches. That’s a compromise that would give pause to those who’d prefer some gold-standard Euro rolling stock purity right out the gate: is it better to get more systems with skin in the electrification game first, even if that means gritting through another generation of heavyweight U.S. procurement?

          Note that there is also great potential for Bombardier to eventually serve up a single-level version of this EMU if the MLV version is successful. It’s just not going to be catered the same way as the NJT + GO orders for trainlining with legacy fleets to defray costs, because no commuter agency has ordered flats in so many years. On the plus side, they’d probably be importing a fresh and more svelte carbody from their Euro offerings since the aluminum Comets/Shoreliners are much too dated to accommodate for backwards compatibility. Problem is they’ll be busy enough with years of bi-level orders that adding a flat version to the lineup probably overshoots the timetable for SEPTA’s Silverliner IV replacement order and MBTA’s first-chance dip into electrification. So SEPTA’s probably still going all-custom for the Silverliner VI, wretched track record be damned. Maybe for the T that means buy NJT-spec bi-level MLV EMU’s first for the Providence crowd-swallowers, grin and bear it for a few years when those have to be used on the intracity Fairmount Line, then differentiate the second follow-on order of EMU stock in the flats packaging after electrification of a few more lines and assign by most appropriate schedules. It would still be a homogenous-propulsion fleet they could bank on for life-of-fleet cost control, but the packaging choice would have to contour to what’s readily available on the market at least modification at the time they make their go-for-it decision on electrification.

    • Alon Levy

      That’s what it’s intended for, but the current trains can only go 160 mph because that’s what they were built for. The new ones can go IIRC 200 mph.

      • adirondacker12800

        Once they come up with regulations for tracks that can serve trains going faster than 160. And upgrade them to that. There’s that problem with constant tension catenary to fix too. Run 180 mph trains between New Brunswick and Trenton you can’t have NJTransit trains clogging up the express tracks at 105. Or long distance trains. One of the secondary reasons why more Regionals don’t stop at places like Princeton Junction is that the local tracks are Class 5 or 6. Or 4. One of the reasons the new long distance cars are good for 125 is so that they don’t clog things up between Philadelphia and Baltimore.
        The foamers love to go on and on about how the Amfleets should be replaced with something that can swap out of service along the NEC to go Denver. Nah, move the Acela IIs to the Keystone and replace the Amfleet/ACS64 combo with Acela IIIs. And upgrade the local tracks in New Jersey and Pennsylvania. And electrify to Springfield and Saratoga Springs. CDOT can then run M12s to Springfield…

        • Alon Levy

          They’re planning to increase Acela speed from 135 mph to 160 in segments south of New York to be fitted with constant-tension catenary. Not sure if they’re also increasing speed from 150 to 160 mph in Rhode Island and Massachusetts.

          • F-Line to Dudley

            All 150 MPH territory in MA and RI was supposed to be raised to 165 MPH when the new 165 territory in Jersey came online, but that was close to 18 months ago and nothing’s yet happened. Don’t think there’s any regulatory snags holding it up. Possible they’re just waiting for ongoing construction to the south to fully wrap so they can lump all time savings into a single Acela schedule revision.

          • adirondacker12800

            The design speed of the current Acelas is 165, almost every operator runs at 90 percent of the design speed. Which would be 148.5. They aren’t going to be running at 165 until the new trains arrive. Wikipedia says the test train should be delivered next year with delivery of the fleet in 2021 and 2022. Wikipedia also says at 160 until there are additional improvements. Like regulations for track with trains running at speeds higher than 160.

          • F-Line to Dudley

            No, ad12800, that is completely incorrect. The current Acela fleet has been certification-tested for 160 MPH service operations, with a differential of 5 MPH off max design speed…not 10%. It’s on the Wikipedia page for Acela Express and direct-quoted in news article after news article about the higher-speed testing if you care to Google it. The formal uprate to 160-in-service is now scheduled to happen in 2020…after 3+ years in delays attributed to behind-schedule work in NJ. This uprate will go live on the CURRENT fleet, not held off another 2 years for the new stock.

            I don’t know if the MA and RI segments are going to get lumped in simultaneous with 160 MPH in NJ, but I presume those 3 segments in tandem net the most meaningful schedule gains so holding the uprates to sync in a single schedule revision makes practical sense.

          • adirondacker12800

            They’ve run current Acela’s through Princeton Junction etc. at and above 165. The the Wikipedia article tends to reference dead links or ancient stuff.

            I think the part of the Wikipedia article you are reading far too much into is ” $450 million was allotted by President Barack Obama’s administration to replace catenary and upgrade signals between Trenton and New Brunswick, which will allow speeds of 160 mph (257 km/h) over a 23 mi (37 km) stretch. The improvements were scheduled to be completed in 2016, but have been delayed; the project is now scheduled to be finished in 2019. ” It doesn’t say shop queen tanks are going be clipping along at 160. Or more than 135 for that matter. I don’t really care all that much because the decision in late 2022 is to make them into reefs, tolerate the shop queenyness and use them for Regionals or Keystone or gut em and make them into cabbage cars. The people paying high prices so they can brag they took Acela ain’t gonna be on shop queen tanks.

          • F-Line to Dudley

            I wasn’t referencing Wikipedia. I was referencing real news articles from Fall 2017/Winter 2018, with quotes from the mouths of Amtrak spokesflaks saying 160 would go online into the schedule in 2020. On scheduled service meaning with current fleet, because new fleet won’t be available in 2020. And 2020 being the MORE up-to-date projection for completion of the delayed NJ project than the dated 2019 Wiki copypasta you’re quoting. Please take one stinking minute of your time to Google for this before launching into the prefab Mad Libs sidebar about shop queen tanks and “the foamers” and blah blah blah.

          • adirondacker12800

            Silly me when you said “It’s on the Wikipedia page for Acela Express” I assumed you meant WIkipedia not some specific source you can cite with a link. Since you weren’t more specific I clicked through a few that seemed likely, they are either old or dead. It’s real easy to cut and paste links in, there’s more than one way to do it but ctrl+c and ctrl+v seems to work most reliably for most people. I really don’t give a flying leap because once the new fleet is running it isn’t gonna matter what they do with the shop queen tanks. If they push the current ones for 15 month big ginormous whooppppppeee! I hope you get car numbers as they whiz past.

  3. fbfree

    On Metra, it should be noted that the Metra Electric, going to Millennium station, is the only Metra line without a severe need for new equipment, having fairly recently updated the entire electric fleet. Rather, it’s electrification and conversion of the existing diesel lines that’s needed (starting probably with the Rock Island, as it’s the only other wholly owned line and relatively isolated.)

    • ardecila

      I like the idea of starting with RI for electrification, if they went with a DC power system like Metra Electric they could potentially interline with the Blue Island Branch of ME. Unfortunately the best lever for demanding Metra electrification is the toxic air in the downtown terminals, and RI’s LaSalle St has probably the best air quality of all the terminals.

      • ardecila

        IMO the ideal first project would be MD-N. Wholly owned by Metra, improves air quality in Union Station, and the greater acceleration from electric rolling stock would compress schedules and create spillover benefits to the popular Hiawatha Service even if Amtrak continues to use diesels. It could also allow for an EMU express service to O’Hare and/or a future crosstown line like the CrossRail Chicago plan.

        • Alon Levy

          Not UP-N? It’s, well, UP, but it has no freight, and it’s a viable express overlay next to the (by far) busiest L line, which is going to be shut down for reconstruction soon anyway…

  4. Max Wyss

    The argument that there is always customization is insofar true that every customer has different ideas about the interior, but more than often, signalling equipment is customized. However, the leading manufacturers have product lines which allow for a relatively easy adaptation. There will be a design phase, but it will be way shorter (and therefore cost less) than for a completely custom-made vehicle. But even with completely custom-made vehicles, an experienced manufacturer will use standardized components.

  5. Oreg

    Yay! 🙂

    About Millennium Station: It’s a terminus with 13 tracks, according to Wikipedia. If all passengers get off at a terminus, bilevel dwell time does not increase the travel time along the line, so they should be fine, no? Also, isn’t the bottleneck at Millennium more the capacity of platforms and exits rather than the train doors? If that is not the case, longer dwell time blocks a platform for longer, reducing the throughput of the station. Is that the issue you had in mind?

    • Alon Levy

      Why did I think Millennium Station had 7 tracks and 4 platforms, not 13 tracks? If it’s 13 tracks as Wikipedia says then definitely bilevels should be fine, provided there’s no through-running.

      The capacity of platform egress points, as at Penn Station, favors single-level trains, because they don’t dump so many people onto one platform at a time.

      • Oreg

        Here’s a track map, photographs and some background on Millennium Station:

        I seems that, unlike Penn Station, the only egress point is the end of the narrow platform. Bilevels have higher capacity per length but fewer doors, so they release more people overall but fewer per time, right? That should help manage platform overcrowding.

        • Alon Levy

          Ah! So it’s just the Metra fief that’s only 7 tracks (5, really). Is there a way to add more turnouts to the interlocking so that trains can more easily connect to the NICTD tracks? With just one turnout, I wouldn’t trust a rush hour Metra Electric train volume to run reliably to those tracks.

          • fbfree

            I believe that the turnout itself is in a covered cut, with few or no columns, where the layout can be easily modified. The layout close to the platforms would be harder to change.

          • adirondacker12800

            Wikipedia says Metra Electric has 34,000 weekday riders and the South Shore has 11,000. They all stuff themselves onto trains so that they arrive between 7:45 and 8:45 that’s 22,500 of them. And 22,500 of them between 16:45 and 17:45 to go home. Probably not a problem. And they don’t stuff themselves all into one hour. Leverage all four tracks of the mainline so there is train every five minutes with 1,000 people on it that’s 24,000 an hour in each direction. Probably not a problem.

          • Alon Levy

            I just had to let two RER B trains go before getting on the third at 6 in the evening, Les Halles -> Gare du Nord. The dwell times were around 90, 90, and 150 seconds. The main takeaways:

            1. French people are savages who don’t take off their bags on crowded trains, don’t consistently let people off the train before getting on, and don’t move into the aisles.
            2. Trains really need to be designed with better crush capacity in mind, which means not obstructing the aisles. American trains are wide enough that 2+2 seating should be fine, which it really isn’t here, but 2+3 leaves aisles that people wouldn’t readily stand in.
            3. Even at absolute crush, the platforms at Les Halles are way wider than necessary, and could most likely be narrowed by a factor of 2. My Lower Manhattan station proposal has 10-meter platforms and I’m more worried about escalator capacity than about platform width.

          • Oreg

            He, I’ve experienced the French as generally inconsiderate—not just in famously rude Paris.

        • fbfree

          The Metra side of the station has egress at both ends of the platforms. The NICTD platforms only have the one exit.

    • ardecila

      Also don’t forget that a non-trivial number of people exit at Van Buren St, which is the more convenient option for jobs south of Monroe St. As usual, ME is the odd duck in Metra’s network with two popular downtown stations instead of one, and (anecdotally) a majority-minority ridership base.

  6. Michael James

    Alon Levy, 2018/11/21 – 16:12
    I just had to let two RER B trains go before getting on the third at 6 in the evening, Les Halles -> Gare du Nord. The dwell times were around 90, 90, and 150 seconds.

    This is your continuing frustration with the single tunnel that RER-B and -D have to share between Les Halles and Gare-du-Nord. I kinda agree, but as before, wonder if it is the priority use of $5bn to build it. OTOH, you might be a slow learner; I mean trying to use the two busiest transit stations in Paris (in fact, Europe) at peak or close to peak! You wouldn’t find me there. I would either walk (it’s 2.2km, 20min probably less than your entry to the station + waiting time + transit time) or use M4 which is nominally slower … except faster than letting two RER trains go past. Or M7 and walk from Poissonniere or Gare-de-l’Est etc. Or Velib it (no, I think I’d give the street traffic at that time a miss unless there is a genuine cyclepath ROW. And of course … weather.)

    Anyway I have a suggestion. Strap on your seat belt because it’s almost Muskian. Not really (not least because Musk took his e-sled idea right off ThyssenKrupp), and I have suggested this before on your blog, for a connection between Central and Penn Stations: travellators between Les Halles and Nord. Unlike the 400m of somewhat grim fluoro-lit pedestrian underground tunnel(s) connecting Chatelet with Les Halles, I reckon do it above ground. Like M2 and M6 but purely pedestrian. Maybe go for the latest ThyssenKrupps maglev version so it can cover the distance at 3x or 4x walking speed, thus reducing it to about 5min, but really 2x and 10min is hardly awful if old-tech is used. Build it in a glass canopy above the street (rue Poissonniere, La Fayette) and it would be quite inconsequential to Parisian iconography–in any case this route is somewhat grotty (the old red light route famous in many roman a Paris-noir). At road width there’d be space for two sidewalks plus three travellators–in contraflow arrangement, ie. two/one for peak/non-peak directions, plus one spare when one fails. [NB, a huge advantage of maglev is that it hardly ever fails.] Give it some greenery and cafes at the junctions of each travellator (probably every 200m?) and voila, you’d have a Promenade Plantée & High Line, or Hong Kong’s Mid-Peaks outdoor-escalator, blessing of ‘cool’. Who wouldn’t prefer to ‘walk’ this link than fight to get on, and then off, crowded transit trains, especially in Les Halles? Sure, not those who are already on the train before these stations but all those who originate nearby (which is a lot at Chatelet-Les-Halles) and maybe quite a few who are transferring from the Metro. It would serve both RER-B and -D travellers plus the suburban trains at Nord, plus l’Est next door (in any case these two stations are being integrated in the massive redevelopment).
    Of course there are several bottlenecks in the transit system (any transit system, like NYC’s Central-Penn) that could use such an alternative that doesn’t cost $5+bn and ten years of disruption. In fact in 2002 Paris tried their high-speed travellators on the Montparnasse—Bienvenüe link, which is a long underground correspondance not unlike Chatelet-LesHalles. Alas it failed because of pedestrians failing to learn how to enter and exit the new changing-speed travellators. Also, it is a straight-line less-than-2.5km between St Lazare and Nord; you could have a travellator junction at La Fayette/ Poissonnier!

    And in time-honoured tradition, there ain’t nothing new under the sun. Look what I found when refreshing my memory:
    How the moving walkway nearly overtook the Metro
    Paul Collins, 06 August 2009
    When Paris hosted the Exposition Universelle in 1900, it unveiled its vision for the future of transport. Below ground, the city’s stylish new Metro made its debut, while above ground was something more avant garde. The trottoir roulant was a moving walkway that circled the fair in a 3-kilometre loop, its articulated wooden segments “gliding around like a wooden serpent with its tail in its mouth”, according to one reporter. Nearly 7 million visitors hopped on. A few even brought folding chairs, which proved useful when one woman gave birth in transit. Her child was promptly christened Trottoir Roulant Benost. A new kind of traveller had been born.

    Nothing new … except technology may have caught up with the concept which was basically a good one.

    • Alon Levy

      It’s 700 million euros, not 4 billion, provided some periodic shutdowns are acceptable (and there already are occasional shutdowns for maintenance, with passengers told to take M4 or various suburban buses to CDG).

      And yeah, the fact that the high-speed travellator didn’t really work at Montparnasse/Bienvenue is a good reason not to try to impose the same system on Central Paris.

      • Michael James

        I assume they are giving priority to the Haussmann-LaDefense tunnel for RER-E, which in turn will bring relief to all the others (i.e western RER-A & D and indirectly to B and to Chatelet-les-Halles itself). Possibly they believe it may make the problem go away, or at least they are going to “wait and see”. It seems a pity they didn’t do that tunnel right at the time (1970s) they were building Les Halles and the one for RER-B. But remember that that was all highly controversial and super-expensive at the time (to this day there are econocrats who think it was all excessive). Perfect transport planning is so easy in retrospect.

        The Montparnasse-Bienvenue travellator was 20 years ago and was completely different technology.

        You know, just like with the American’s long-delayed reform of their FRA, you should be wary of closing your mind to new-tech developments, even of old concepts. I still think what I proposed some time ago, that you should do a review of maglev in its various forms, HSR, Metro-transit (as in Japan, China & S. Korea) and now these new developments by Thyssen-Krupps (elevators and travellators). It is not at all comparable to Musk’s fantasies because working examples exist for them all.

        • Michael James

          Magnetic levitation: the return of transport’s great ‘what if?’
          34 years after the first commercial maglev line opened in Birmingham, the futuristic tech enjoys an Asia-driven rebirth
          Christopher Beanland, 27 Nov 2018

          Lower-speed maglevs include the new Beijing line, an 8km route in Qingyuan scheduled to open in September and others in Changsha and Seoul. Dozens of Chinese cities – including Chengdu, Shenzhen, Wuhan and Guangzhou – are considering new maglevs. Systems have been proposed recently in Bangalore, Hong Kong and Taipei.

          “For decades maglev couldn’t offer a better product in terms of price and efficiency than high-speed rail,” says Smirnov, an engineer himself. “Now, thanks to the work of many scientists and engineers, maglev offers speeds much higher than trains with wheels and has the lowest operation costs of any transport.

          This is a report on that St Petersburg conference on maglev held this month. So this might just be the usual boosterism such conferences produce … Need more info than a Guardian article.

    • Henry

      It is worth noting that the Central Mid Level escalator is considered, for all intents and purposes, a white elephant that never resulted in traffic relief. It’s just a vertical toy streetcar.

      • Michael James

        Henry, 2018/11/27 – 00:42
        It is worth noting that the Central Mid Level escalator is considered, for all intents and purposes, a white elephant that never resulted in traffic relief. It’s just a vertical toy streetcar.

        Traffic relief? HK roads are saturated and so it is impossible to reduce congestion (any reduction is like a vacuum and immediately filled) so no traffic engineer would think such a thing. Perhaps they had to try to fool some fatcat on some committee that his daily ride would be easier if they could get the plebs onto the escalator? But if you think 85,000 people using it every day to traverse the 135m vertical climb on mid-levels a “white elephant” then you have a different definition to mine. By my rough calcs in another decade more than a billion people will have used it. Naturally some of the uber-rich who live up there probably would use their chauffeured Rolls no matter what (HK has the highest number of Rolls Royces per capita in the world).
        And it is such a disappointment that they are currently refurbishing it (it’s 26 years old). And then buiilding another 230 …
        Hong Kong to become ‘city of lifts’ to aid aging population
        By Vanessa Ko | September 13, 2012.

        … a high-profile announcement that Hong Kong will be installing 230 outdoor elevators over the coming years …

        But the fact is such a project could make a big difference in the mountainous territory for the growing population of elderly people, as well as others who have trouble getting around. Anyone who wants to see an elevator installed somewhere can send in a request.
        The government expects to spend $12.5 million on initial planning, and then $125 million per year to build the facilities.

  7. Nathanael


    The first procurements which this *should* affect are Amtrak, who needs to buy new equipment, like, yesterday. Cross fingers, Amtrak should recognize the benefits and ask for CEM-based designs.

    VIA is actually even more badly in need of new trains, but has no money at all.

    I hope the MBTA is next after Amtrak.

    • Nathanael

      Even for diesel-hauled coaches, this is going to reduce fuel usage substantially. And probably improve acceleration, for long trains at least. But most critically, it’s going to massively cut the cost of acquisition of coaches.

    • Nathanael

      And I discover that VIA got money in the last Canadian federal budget, so they’re going to be the first to benefit. Their new trains will be from Siemens.

  8. adirondacker12800

    Steel, even stainless steel, is relatively cheap. And the cost per passenger or per meter isn’t all that outrageous for the stuff they are buying now.

  9. ardecila

    What are the practical implications of this new regulation for traditional push/pull diesel operations? It’s so wonderful that we are getting this new regulation only AFTER Amtrak has already ordered new locos and coaches for its popular regional Midwest and West Coast routes.

  10. Patrick

    The LIRR M-9 procurement is broken up into two major parts, so they might be able to control the damage if they play their cards right. There was always plans to have two separate types of M-9’s, the main M-9 order paid with local funds, and a smaller, separate M-9A order funded with federal funds through the Rolling Stock Liability Reserve for East Side Access.

    The M-9 base order contract for 92 cars was awarded in September 2013, and the contract originally had options for 88 cars for LIRR (M-3 replacement option), 76 cars for LIRR (ESA option), and 155 cars for Metro-North. LIRR exercised the first option and portions of the second for a total of 202 cars.

    The M-9A’s are currently in the RFP process, with the contract award not expected until Q2 2019. The timing is such that they could go an entirely different route and seek alternative compliance for the M-9A’s if they wanted. That order is currently set for the original 160 rolling stock liability reserve cars + 54 previously M-9 option cars + 30 LIRR option coaches + 200 option cars for Metro-North. The LIRR could conceivably try to worm out of the already exercised M-9 options to put more cars towards a modern M-9A design, but we will have to see how they play it…

    • adirondacker12800

      How many years would that delay the procurement? Changing the specifications this late in the game?

      • Patrick

        Given the poor reliability of the M-3 fleet, the base order M-9 cars should be sufficient to replace that fleet. Other than that, cars are not immediately needed until East Side Access opens, in four to five years. That should be enough time to go through a new car procurement…especially if the idea is to lightly modify an existing European MU car, not starting from scratch.

        • adirondacker12800

          I’ll bite, what are you going to “lightly modify” so it can platform in Grand Central or Penn Station. Why is that so much easier than having Kawasaki or Bombardier or Alstom or Siemens or …modify what they make for the North American market? Why would they want to import stuff all the way across oceans when they can get parts in North America and have assembly plants here?

  11. Josh

    Does this interface at all with the PTC requirements, and the likelihood of them being done by the end of 2020, or around when new trains would come into service under the new standard (after the procurement process)? As in, reduced crash probability decreases the need for crash protection.

    • Alon Levy

      The FRA rejected the notion that PTC obviates the need for crash protection, on the grounds that PTC protection is not absolute. I looked in the document and didn’t see any evidence of retooling, though: CEM is still designed around the sort of train-on-train crashes that PTC is supposed to protect against, rather than around (say) grade crossing accidents.

  12. Henry

    Does this affect DMUs as well? As far as I remember there is only one FRA-compliant DMU being built (for SMART) and those have had issues.

    While electrification is the best, realistically speaking DMUs would enable lots of small, cheap projects to be feasible, and Europe has no shortage of DMUs being sold.

      • Henry

        Wouldn’t that be the same for EMUs, though? No off-the-shelf European trains are manufactured in the US today, electric or diesel.
        IIRC Stadler has also managed to get FLIRTs (TexRail) and KISSes (Caltrain) into the country as well.

        • F-Line to Dudley

          EMU’s are usually ordered in much larger numbers than DMU’s, so the purchase scale is night-and-day. SMART’s first order of Nippon-Sharyos was 14 cars, UP Express was 18 cars. For Stadler GTW’s it’s 20 for NJT RiverLINE, 6 for Capital Metrorail, 11 for A-Train, 8 for eBART. O-Train has 12 Alstom Coradias, Sprinter runs 12 Siemens Desiros, WES runs 3 Colorado Railcars.

          The *smallest* EMU orders of the last 25 years were: NITCD with 10 single-level Nippon-Sharyos; NITCD with 14 Highliner II’s; Metra Electric with their first 25-unit batch of Highliner II’s in 2005 (since supplemented by 159 more). Nobody else has ordered fewer than 48 at a time. LIRR, MNRR, Metra Electric, and SEPTA have all placed single EMU orders larger than the entire combined continental installed base of DMU’s, and NJT is about to do the same.

          It’s going to be awfully hard for the DMU market to take off at all when the orders are so tiny. Then feed it through Buy America where assembly has to be laundered through U.S. plants only a few units at a time, and most builders can’t make any money selling them here. Nippon-Sharyo, especially after all the micromanaging SMART put them through, seems to have no interest left in peddling those vehicles. Stadler’s still doing OK, but who knows if the GTW is close enough to FRA compliance under the new regs to bust out of its time-separation ops niche. Unless the Chinese are hungry and willing to write off large losses peddling DMU’s as a gateway to larger/more lucrative contracts, the market for DMU’s is looking like it’ll be staying pretty ice cold for awhile.

          • Henry

            All you need is one order; the Boston Orange Line wasn’t very big, but was enough to convince CRCC to open a factory. Same with failed Wisconsin higher speed rail and Talgo.

            Big diesel renewals are coming. Indigo Line is shovel ready for a new MBTA administration. MetroNorth and LIRR diesels are essentially EOL. DMUs in particular would be a massive improvement in LIRR diesel territory since the current diesels are so bad.

          • Alon Levy

            Yes, and CRRC took a loss (or at least the paranoids in Congress think it took a loss) while still bidding near the high end of international single-level EMU costs, which are in the $80,000-120,000/linear meter range.

          • F-Line to Dudley

            Sorry, Henry…none of those agencies are buying DMU’s.

            LIRR and MNRR are partnering up on common diesel fleets: MLV-dimension bi-levels to replace the LIRR C3’s and MNRR’s Shoreliner and Comet flats. The MNRR P32AC-DM and LIRR DM30AC dual-modes are going to be replaced in a triple-agency procurement with NYSDOT that also replaces Amtrak’s P32AC-DM’s. Supposedly that one is Siemens’ to lose since they’ve pitched a dual-mode version of the Charger. And if the dual is going to be a Charger, then LIRR can either replace its DE30AC straight diesel with straight Chargers on the same contract or buy additional units of the dual with or without the E-mode compartment filled. As much as DMU’s would help some ops, right now they’re still prioritizing one-size-fits-all fleets that can be sent to Penn Station if needed, and can run on crowd-swallower schedules if needed. They’ve already announced plans to electrify the Central Branch connector at long last, and if they finally electrify to Port Jefferson diesel territory will be limited to just a few Patchogue short-turns and the east end + Oyster Bay Scoots. Those moves shed enough of the mixed-load usage of the diesel sets to the point where you can target just the Scoots for a right-sized fleet. But they’re not there yet, and the next procurements are probably going to happen before they can commit funds to Port Jeff electrics.

            The MBTA’s DMU purchase offer was for only 18 units off SMART’s options, and a fleet that tiny just didn’t make a lot of sense especially when it required putting up with all the problematic overcustomization cruft that SMART ruined those vehicles with. Now they’re doing what they should’ve done all along and are talking electrification of the Fairmount Line and lumping an EMU order in with the huge Providence Line fleet requirements for real ordering scale. It’s not clear Nippon-Sharyo would even bid their DMU again with all the corporate turmoil they’re in after effing up Amtrak’s Midwest/Cali bi-level coaches to the point of losing the contract. And because of that it’s entirely possible that the first northside “Indigo” routes turning at Route 128 are going to have to make do with loco-hauled 3-4 car sets of single-level coaches for service starts. Electrification up north is going to trail electrification on the southside by many years since there’s no preexisting physical plant to build off of, and if the DMU market is going to be this moribund it’s better to at least get the service levels seeded sooner with the best the current equipment can do rather than do nothing at all until the DMU market wakes up. Sub-optimal enough to never be intended as a permanent condition, but as a temporary bridge era some precision-configuration push-pull consists on the inner shuttle routes is a lot better than nothing. And domestic DMU’s are almost literally a nothing market going into 2019.

  13. Pingback: What the Latest Passenger Rail News Could Mean for Cascadia
  14. Pingback: Why American Costs Are So High (Work-in-Progress) | Pedestrian Observations
  15. Pingback: The NTSB Wants American Trains to Be Less Safe | Pedestrian Observations
  16. Pingback: The Private Sector’s Role in Transit Innovation | Pedestrian Observations
  17. Pingback: Metro-North Doesn’t Know Best Industry Practices | Pedestrian Observations
  18. Pingback: New York Rolling Stock Costs are Skyrocketing | Pedestrian Observations
  19. Pingback: MAGA Trains | Pedestrian Observations
  20. Pingback: How Should LA Urbanize like East Asia? Mass Transit! | Entrepot
  21. Pingback: The Need to Remove Bad Management | Pedestrian Observations

Leave a Reply

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

You are commenting using your 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.