20 September 2016

EMU Brochure

Credit: Stadler Rail Group
This week is InnoTrans 2016, the world's biggest rail industry trade show, held in Berlin, Germany. Stadler has a big presence there, and rolled out an updated website and downloadable brochures. Among them is a preliminary brochure for Caltrain's new KISS EMU, affording the first public glimpse into design details of Caltrain's new fleet.

Key Specs

The Stadler KISS (recently re-branded KISS160) as configured for Caltrain is a high-performance train that will have no issues sharing the corridor with high-speed rail.
  • Power: 6000 kW (8000+ hp)
  • Acceleration: 1 m/s2 from 540 kN (implying a loaded mass of 540 metric tons)
  • Braking: rated at 8000 kW, with most of the energy fed back into the electric grid
  • Vehicle width: 3 meters (comfortably inside US loading gauge)
  • Maximum speed: 110 mph
  • Straight sides, taking full advantage of US loading gauge, unlike European KISS
Train Layout

The six-car EMU shown in the brochure features dual boarding height doors throughout, to share platforms with high-speed rail and to enable Caltrain to transition gradually to gap-free level boarding, to cut down dwell times for the blended system.  The train consists of the following:

ParameterCar 1Car 2Car 3Car 4Car 5Car 6Total
Car typeCabMiddleMiddleMiddleMiddleCab-
Number of powered axles24040212
Seats, lower level3823638638149
Seats, upper level525260526052328
Seats, intermediate level10+21010+161010+1610+296
Seats, total102859210092102573
Bike spaces--40-40-80

As delivered, the upper level boarding doors are sealed and 5 seats are configured longitudinally in each vestibule on the intermediate level, accounting for 10 seats per car. These seats will be temporarily removed during the transition to level boarding with high platforms (when both sets of doors must be cleared of seats), diminishing seating capacity by roughly 10%. Once Caltrain achieves 100% level boarding, that seating can be restored on the lower level of each car.

The layout for all six cars can be viewed by zooming into the original PDF brochure, or more conveniently as individual graphics extracted below.


The seating diagrams reveal that the ADA bathroom displaces 15 seats.

Interestingly, the Stadler diagrams show two wheelchair spaces on the lower level of every car, which appears to imply that ADA wheelchair lifts would be provided in every car to cross between the lower and intermediate levels of each car.

Future Capacity

Caltrain's order from Stadler includes options for another 96 cars, which are planned to be exercised to expand the fleet to 24 trains and lengthen all trains to 8 cars or about 200 m long.

With the seating layout and train car dimensions defined, it is possible to predict the seating capacity of an 8-car train. The basic 6-car train already has all the traction equipment, so the two extra cars would likely be trailer cars similar to cars 3 and 5 above, except without the bike storage area on the lower level. Each of these extra cars would have 60 seats on the upper level, 46 seats on the lower level, and 26 seats on the intermediate level, for a total of 264 added seats. This would bring each train's ultimate capacity to 837 seats and 80 bikes, with plenty of additional space for standees.


  1. All very nice I would say, except how is it going to be funded? Certainly not under the present CalTrain funding scheme, which included over $800 million from High Speed Rail funds, which will never pass the courts, when litigation begins.

  2. Neat, glad to get an early look at these!

    Question, though - are these going to have a lower seated capacity than the existing trainsets? There's a good chance that my memory is off, but I thought that the existing trainsets were already well over 600 seats.

    1. Unequivocally YES the new EMU fleet will have fewer seats per car than the current locomotive-hauled fleet. How embarrassing!

      Some ascribe this to the dual boarding heights, but a careful look at Stadler's layout reveals the true culprit: there is an entire locomotive's worth of traction equipment stuffed in all those cabinets marked with an electric flash symbol. It's hard to hide a locomotive inside the train without displacing a large number of seats.

      Caltrain should have gone five-abreast and I think they will regret not doing so. Maybe there's still time, given Stadler's demonstrated willingness to tinker with car shell dimensions!

    2. which is still weird to me that the electrification project specified a loading gauge capable of running trains as wide as 3.4m, but EMU RFP specified for narrower 3.2m trains, with the Stadler bidding even narrower 3.0m trains. Cost saving, perhaps?

    3. Actually, there are two locomotives worth of electric equipment stuffed into the trains.

      Other comments: For a KISS, the trains are on the lower power rating, considering the 1m/s^2 acceleration (compared to the bls Mutz with 1.3 m/s^2). Considering that, I wonder whether the optional additional cars won't come along with one powered truck…

    4. Has the HSR project settled on a loading gauge yet?

    5. @Joey: CHSRA more or less decided on 3.4m wide loading gauge: http://www.hsr.ca.gov/docs/programs/eir_memos/Proj_Guidelines_TM600_01R0.pdf

    6. @Max: OK, so if we suppose one of the two cars to be added is like car 4 and the other like car 5 without bikes, we get another 100 + 132 seats or 805 seats per 8-car consist, and increase power to 8 MW.

      @Joey: HSR has a composite loading gauge defined in their tech memos that encompasses all the possibilities, including 3.4 m wide Shinkansen. Presumably they had a hand in defining the new platform interface specified by Caltrain for the high doors, 50" above top of rail by 72" from track center. That is plenty of clearance for a 3.4 m wide train.

    7. It is eyebrow raising that they didn't use the entire loading gauge then. Sure it adds a bit of weight and cost, but even if you don't add a fifth seat the extra aisle space seems worth having.

    8. The Bombardier Bi-Level car design has much more efficient use of space. Just consider how cleverly the ADA bathroom is tucked away on those trains compared to with the KISS. For fun, played around with diagrams and came up with this:

      * KISS cars with bathrooms lose space for 10 seats.
      * Bombardier loses 3 seats due to efficient tucking of it.

      Shorter upper deck:
      * Bombardier stuffs 72 seats upstairs
      * KISS has 60 seats upstairs.
      The upper deck is overall shorter to the tune of losing about 3 rows ~ 12 seats

      Without a blueprint diagram, it's unclear why the upper deck is shorter especially on the unpowered cars. Perhaps they need to keep body shells all identical to simplify manufacturing?

      Once caltrain if fully high platform, removing the 4 lower doors should permit adding about 4 seats per door or 16 seats per car... An extra 96 seats per 6 car train gets us to the equivalent capacity of a 5-car bombardier set.

    9. As to the Moscow trains from the airports to the city centers, there are longer cars in the center of the 6-car ones. Here's a link https://wwwstadlerrailcom-live-01e96f7.s3-eu-central-1.amazonaws.com/filer_public/9c/57/9c57a2f9-9ff3-4c1c-aeb8-ac607eee4ba0/kissae0513_en.pdf

    10. Interesting. In the 6-car set, cars 1, 3, 4, 6 are longer, while 2 and 5 are shorter. It's also the short cars that have electrical equipment. If I had to guess, it might be due to extra weight of transformers.

    11. Yeah, the extra weight of the transformers must be it… The thing is that the Aeroexpress trains do not have transformers, because they run under 3000 VDC…

      To me, it looks as if the main reason for the shorter cars were to stay within the (roughly) 150 m length (6 car units) or 100 m length (4 car units).

  3. As predicted, the ADA access is horrible. For high-level platforms, wheelchair lifts will be needed to get on and off the train. It would actually be preferable for Caltrain to not build any high platforms, leaving the 8" platforms as-is.

    1. I don't follow, because 8" platforms won't need wheelchair lifts to get on and off the train?

      Surely that's not what you meant.

      The 8" lift must be done (twice!) during station dwell, while the 50" lift can be done while the vehicle is in motion. That saves a lot of time & hassle for everyone, and ensures short and predictable dwell times for a reliable blended system. Not sure why this is so hard to grasp.

    2. Existing mini-high platforms can still be used with the lower 22" doors until the platform height is raised.

    3. "Existing mini-high platforms can still be used with the lower 22" doors until the platform height is raised."
      Way to go with the two billion dollar "investment" in "improved" Caltrain "service"!

      Mini-highs until 2040! Or later. Much, much later.

      This is going to be great!

    4. Clem,
      So after $2 billion "investment", we go from requiring 2 lifts to...2 lifts. All that's been done is to move the lift to the inside of the train. And for bikes, all that's been done is to move the stairs from outside to inside the train.

      The situation might have been different if some bikes and wheelchairs could be stored in the vestibule -- but the pictures don't seem to show this possibility. And some of the cars have too much electrical equipment in the way.

    5. All that is only deplorable if you totally ignore the issue of HSR compatibility, or if you believe that the unicorn of low-floor HSR actually exists.

      The Stadler EC250 (remember, the one with 550 mm boarding?) has a "low" floor height of 940 mm, with a step up to a 1200 mm gangway between cars. Videos of the interior are starting to show up on Youtube from InnoTrans, and I think it's fair to say that the idea of a low-floor high-speed EMU is nowhere near being realized. That being said, it's a darn nice train.

      What we get with the Caltrain KISS is a train that allows a gradual transition to level boarding over time (another big logistical and financial puzzle), which in turn enables short and predictable dwell times, which in turn allows more trains to run reliably on the corridor with less timetable padding--something that will be critical when HSR starts sucking up line capacity. We also get shared platforms at Transbay, Millbrae, Redwood City and San Jose, avoiding easily a billion of inefficient HSR-dedicated station infrastructure. No more Diridon Pan-Galactic in the sky.

      Still, one can ignore all these things and complain about bike access.

    6. TIL that the TGV duplex is vapourware....

      BTW, can you provide links to these Innotrans Youtube videos? According to this article posted today, the EC250 will have "step-free" access for both 55cm and 76cm.

      In this youtube video from 2 days ago, the Stadler CEO says "We have always promoted stepless entry in our trains. Step in, not step up is our motto. So we decided to do it with high-speed train as well."

    7. That means, that the entrance area is 55 cm (standard in Switzerland (and I think, also Italy)) or 76 cm (standard in Germany) above rails. Within the car, there will be ramps (or steps) to get to the regular floor heights.

    8. Max,
      I'm not seeing any steps/ramps between the door and main floor area, at least where wheelchairs ride: https://www.zwomp.de/2014/05/20/nose-ec250-sbb/ec250_prm. Existing Stadler low-floor rolling stock is very accessible, and based on youtube videos (and CEO comments), the EC250 seems to have similar layout.

    9. Please, enough with the TGV Duplex already. The TGV Duplex cannot reasonably be made ADA compliant due to the need for interior lifts to reach the lower deck level @ 320 mm, and ADA bathrooms in every car. The TGV Duplex also represents vendor lock-in for a design that represents a minuscule portion of the worldwide supply of HSR train sets, some 90% of which is high-floor single level with no visible signs of market evolution towards bilevel or low-floor solutions. Finally, the TGV Duplex cannot be made as an EMU (the electrical bits are too heavy and the mass budget for the double-deck cars is already too tight), and as a loco-hauled trainset has a marginal adhesion factor to operate on very long 3.5% grades, the likes of which do not exist anywhere in Europe. Fun fact: the seat shells in the TGV Duplex are made of magnesium to save weight, because even aluminum is too heavy.

      Here is a video of interior steps between each car in the Stadler EC250. The brochure is on Stadler's website.

    10. By looking at the videos, the low floor level extends up to the second to last set of seats, which is reached with a ramp; then, there is a step up to the gangway area (over the Jacobs bogies). (note that the Swiss / European accessibility regulations are somewhat less strict than the USAn ADA regulations, when it comes to circulation within the train.

      This is pretty much similar to the FLIRT configuration.

    11. In EC250, you can see a small ramp + step. Something that someday might just be a long ramp.

      BTW, I was in Italy a month ago and saw many stations getting converted to level boarding. At the old stations like in Rome, the extra height added results in extra railings for ramps on platforms for the outer tracks even though they're only adding a few inches. Expect to see tons of railings in places like SB Mountain View platform whenever level boarding appears.

      Thanks for clarifying to the few remaining bits about TGV Duplex... it's always been on my mind.

  4. Looks like overall, these will have about the same seating capacity (if bike spaces are replaced with seats) per 6 car train as the single-level M8s do, except these have 2+2 seating while M8s have 2+3. The M8s do have more bathrooms, so it seems like the main thing being gained by sticking to bilevels is 2+2 seating and bilevel boarding.

    1. The comparison with the M8 is an interesting one.

      Metro North 6-car M8: 4800 kW, 633 seats, 155.4 m, 393 t
      Caltrain 6-car KISS: 6000 kW, 573 seats, 157.1 m, probably around 340 t (empty)

      seats/length: M8 4.1 seats/m, KISS 3.6 seats/m
      weight/seat: M8 620 kg/seat, KISS 590 kg/seat
      power/weight: M8 12 kW/t, KISS 18 kW/t

      By the way, the AW3 load referenced in the brochure assumes full seated load, 3 crew persons, plus 6 standees per m2 of suitable standing space (over 800 standees). In other words a crush load of more than 1400 passengers or about 100 tons. This is probably where a major difference lies: it will be difficult to stuff 800 standees in the narrow aisles of a six-car M8.

      The M8 is 3.2 m wide, so I don't understand why the Caltrain KISS can't be made another 20 cm wider to go 5-abreast. It would help Caltrain's case in securing $647M of FTA "core capacity" funding-- it's kind of hard to argue for that when your seating capacity is actually going to drop! Looking at the seating diagrams above, a fifth seat is worth another 100 seats for a six-car train.

    2. It's the damned if you do, damned if you don't situation- if you install 2+3 seating, sensitive Bay Area snowflakes will complain about having to sit in the middle seat, or the middle seat will be used as a place to dump their bags while grandma stands. OTOH with 2+2, during rush hour, people will be complaining about "cattlecar conditions" while moaning about their poor feetsees and lack of seats(!)

  5. I don't see seating capacity argument with BART's new car. Caltrain should follow the same path of BART, meaning ridership demand requests more "capacity" including standing.
    With this opportunity, we hope Caltrain to become "rapid transit" with useful train frequency 7 days a week. So, I was disappointed when they give up option of 3-car short train consist.
    Lower operating cost (one conductor) is need for increasing frequency with local train, late night and weekend.

    1. Configuring some of their option cars as 4-car EMUs, to be doubled-up to 8 cars during the rush, would be a fine way to run more frequency off-peak without operating costs going through the roof.

      That funky automatic coupler you see in the rendering is fully automatic and performs electrical and pneumatic connections without the need for extensive labor in making or breaking train formations. That is precisely its purpose.

    2. 4-car units would indeed be a good set up, if the maximum train length would be 300 m. This would allow 1, 2, or 3 car consists. Apparently, Caltrain does not have that "luxury". So, their current option is most likely the best compromise.

      In most places where KISS are operating in regional/suburban service, there is exactly one person staff on the train… the driver … no need for conductors.

      About the automatic coupler… On the Zürich S-Bahn network, they strengthen/weaken trains normally when changing direction, and that may have to happen within 5 minutes. Because the trains are autonomous units, brake tests are simplified, which gains quite a bit of time (and besides, there is not much to test with disk brakes, except looking at the indicators on the side of the cars).

    3. I don't think the automatic couplers on Caltrain KISS makes much sense, as they will be rarely used. The minimal train length that would utilize the automatic couplers is 2 X 515.3 ft, only San Francisco and San Jose currently have platforms that can support this. So AAR-H couplers more sense from the view of rescue operations, and it would eliminate the need for coupler adapter.

  6. @Clem -- others.

    Caltrain submitted to the FTA this document as part of their process in hoping to obtain the $647 million 5309 grant.

    see: https://drive.google.com/open?id=0B9m407yyFerMQW5KWlNtallPUkk

    On page 5 we read from the table:

    69,000 weekday ridership in 2020
    (The 2020 ridership requires 38 train trip and 6 trains / hour at peak.

    expanding to 111,000 weekday ridership in 2040.
    This would seem to extrapolate to 10 trains / hour at peak in 2040.

    So this brings up the question that as I understand using the 2 track with Passing tracks the capacity of the corridor is 10 trains at peak, yet Caltrain
    is committed to having HSR use 4 trains per hour.

    So the question is, how is this going to fit? Does this mean by 2040 4 tracks will be needed as a possible option?

    1. I think it means they plan to operate 8-car trains with SRO crush loads. An 8-car KISS will carry 800 seated passengers and another 1100 standees, or nearly two thousand people. Because the average trip length is only a bit over 20 miles, a single rush-hour train could serve more than 3000 passengers. Multiply by 6 trains/hour, 6 hours/day of rush (i.e. 36 peak trains per day in each direction, or 72 peak trains per day) and you're in the vicinity of 200,000 weekday passengers. So the figure of 111,000 weekday passengers seems easily attainable without 4 tracks.

  7. So we get UIC EMUs, we get a blended system, gradual transition to high platforms... it seems the design of the system improved a lot from when you started this blog. Did you think this was going to happen? How much influence do you think your writing had?

  8. Something to watch with UK's HS2 efforts. Since HS2 is build to European gauge instead of the small British gauge, double decker trains are possible. Alstom is pitching just that:

    There's also this Aeroliner 3000 being proposed:

    I'm not sure how strict the ADA rules are in UK, but will be interesting to see how that's addressed.

    One idea that's on the Aeroliner 3000, is the first/last "locomotive" cars are single deck EMUs, so conceivably, they could be the dedicated ADA areas.

  9. Indeed it is interesting to compare against HS2 -- because California and the British seem to be making opposite decisions about the platforms. HS2 will be TSI-compliant low floor, and the HS2 platforms at Euston (London terminus) will not be shared with commuter trains, even though it is a very busy station.