Rolling stock for Regional Rail: What and how to buy

If you’ve been reading these posts at all in the past weeks, you’ll have noted that I’m very high on the JKOY Class Sm5 electric multiple unit trainset (pictured above if you’re using a desktop browser) used on Finnish railways for the Helsinki region commuter rail service. I think that Massachusetts should be buying basically this train to replace the diesel locomotives and unpowered bilevel coaches in the current Commuter Rail service. But you can’t sole-source a $2.4 billion government procurement, especially if you have some hope of getting matching funds from the Federal Transit Administration. (Well, you can, but it stinks of corruption and you’d have to be a well-connected Republican and Friend of The Donald to get away with it.) So I wanted to spend some time articulating (no pun intended) what makes this train the right one, in a way that doesn’t explicitly say “it will be a 75 meter electric FLIRT from Stadler” so that other manufacturers would have the opportunity to bid. I’ll also say a few words about a different sort of electric multiple-unit train that I think is Bad and Should Not Be Bought, at least not for the sort of service Regional Rail envisions.

First, though, a long digression about the situation in Finland. When the railroads in Finland were first constructed in the 19th century, Finland was a Grand Duchy in the Russian Empire, and the railroads were built to Russian standards with mainly Russian equipment. This means that the track gauge is Russian, not Standard Gauge as is used on most railroads in North America (and Europe, for that matter). After Finland declared independence in 1917, railway equipment standards in the two countries drifted apart, especially in the modern days of electrification and automatic train control, but they are still compatible if equipment is built to wide enough tolerances. (For most of the 20th century, this served as a trade barrier: Finland was often considered too small a market for European builders, and the Soviet Union did not want to use its hard currency reserves to buy capitalist rolling stock when it could build its own.) However, the Finnish loading gauge — the width of each train at platform level — is almost exactly the same as ours, and wider than the usual European standard, which means that the seating layouts and accessibility could potentially transfer right over.

Prior to Finland’s accession to the European Union in the 1990s, the railway infrastructure, rolling stock, and standards-making were all in the hands of a government department, Valtion Rautatiet (VR, “State Railways”), which had a legal monopoly on all rail transport. In preparation for EU membership, the Finnish government enacted a wide-ranging reform of state-owned enterprises, moving standards, trackage, and rights of way into a new Transport Agency (Liikennevirasto) and creating a new corporate status for VR as VR-yhtymä (“VR Group”), but until 2018 VR maintained its legal monopoly on rail transport. All third-party railroads wishing to operate service in or into Finland could do so only by forming a joint venture with VR, and as of 2017 when I visited, there were only two such joint ventures, one with Russian State Railways to operate passenger services to St. Petersburg (Karelian Trains), and one with the Helsinki regional transport agency to operate Helsinki’s commuter rail network (JKOY). That latter venture is the entity that purchased the class Sm5 rolling stock, starting in 2008; 75 of them had been delivered, in two orders, by the 2015 opening of the Airport Ring Line. They primarily operate on short-distance commuter routes, with longer-distance routes using older VR equipment. In accordance with EU law, JKOY held a tender, and Stadler proposed a cold-weather version of their FLIRT system, already in service in Stadler’s home market of Switzerland and elsewhere in Europe, but this was the first non-Standard-Gauge FLIRT sale; when it entered service it was assigned the class number Sm5. All of the units are owned by JKOY, leased to the regional transportation agency HSL, and operated under contract with VR by VR’s unionized train operators.

Helsinki is a maritime city, but both of the major lines radiating out of the city (the Coast Line to the west and the Main Line to the north and east) run well inland of the Baltic Sea in the parts of the region served by commuter trains. (The Helsinki Metro runs much closer to the sea, and indeed under it in places.) As I noted, the longer-distance commuter trains are operated with VR-owned rolling stock like the class Sm4 EMU, which is somewhat less rapid-transit-like and is more comfortable for the middle-distance journeys some of the express trains make from as far as Finland’s third city, Tampere, 100 miles distant. While Helsinki proper is very dense, especially the districts immediately surrounding the CBD, it is surrounded by large swaths of relatively car-oriented (by European standards) suburbia, and the population density of the entire metropolitan area is not hugely different from the Boston Metropolitan Statistical Area’s — although the absolute population is about a quarter of ours. It’s snowy, cold, and dark in the winter, even more so than here. It’s plenty sunny, but not very warm, in the summer: this means that the HVAC system of the Sm5 is obviously not a perfect fit for Boston: we need less heating and more cooling, but at least we know that these trains are perfectly capable of operating at design capacity even in the winter, with proper maintenance.

(Helsinki even has its own version of North-South Rail Link, with similar financial obstacles; the Urban Rail Loop would be a large, underground “balloon loop” serving three stations in the CBD, allowing trains to enter the city on the Coast Line, serve major employment districts, and then continue without reversing onto the Main Line. If implemented, the I and P trains on the Airport Ring Line would then become true counter-rotating loop routes. While Helsinki has only one Central Railway Station, it is a stub-end terminal and all trains must reverse to continue service.)

Now back to the USA. What are the features of this particular class of train that I think make it close to ideal for Regional Rail in Boston? Here’s a list, with some justifications.

  1. It’s part of a system with broad international adoption, including by other commuter railroads in the United States. Unlike some competing builders, Stadler would not have to build a final assembly plant in the U.S. to qualify for “buy America” because they already have one in Utah to serve their customers here, notably TEXRail in Fort Worth (which is buying a diesel version that has almost exactly the size and interior coniguration we’re looking for). This means that they’ve also been through the Federal Rail Administration’s qualification process for rolling stock that shares track with freight trains — although, thanks in part to acquisitions, the same applies to Stadler’s two biggest competitors, Bombardier and Alstom, as well as the Chinese company, CRRC, which is assembling all of the MBTA’s new rapid-transit cars in Springfield.
  2. They are fully accessible, 100% level boarding, barrier-free trains. This not only respects the civil rights of our mobility-impaired neighbors, it also speeds boarding dramatically, because there are no steps to negotiate, passengers do not need to be concerned about boarding by a specific door or in a specific coach, and passengers can move directly from the wide doors through the aisles into any available seats.
  3. The capacity is right. The existing MBTA bilevel coaches seat about 180 passengers, and single-level coaches seat about 120. The Sm5, which includes an accessible bathroom and ample space for standees and wheeled conveyances as wheelchairs, strollers, and bicycles, has 232 fixed seats in a “3 + 2” configuration and an additional 28 folding single seats which can be used when the space is not occupied by standees or wheelchairs, for a total capacity of 260 seated passengers. In my personal experience, there is plenty of room for standees in addition to the official folding-seat areas, and under load in more rapid-transit-like situations, like the Framingham Line between Auburndale and Back Bay, they could easily accommodate another 50 or even 100 standees. (TEXRail is buying basically this train with a diesel engine stuck in the middle, and they are claiming 230 seated and 217 standees.) That means that, with the sorts of frequency improvements Regional Rail is all about, most trains should be well below the maximum load with only two trainsets in a consist, and all off-peak and weekend service can be accommodated with single-trainset consists.
  4. The physical size is right. These trainsets are 75 meters long, which works out to just under 250 feet. The MBTA’s standard platform length is 800 feet, meaning that for the most crowded runs you could accommodate three Sm5’s on a platform with plenty of slop for inaccurate stopping or slippery rail conditions. A standard Amtrak platform is 1050 feet long, which means that for the initial implementation along the Providence Line you could even run four-unit consists for expresses serving Providence, Route 128, Back Bay, and South Station while finishing high platform construction, performing acceptance testing, working out the schedules, and training operators on the equipment.
  5. It’s ready for AFC 2.0 and conductorless operation. Regional Rail assumes full high-level platforms at all stations, meaning train doors can be released by the operator, rather than forcing passengers to wait for a conductor to drop the traps and open doors manually. As presently operated in Helsinki, there are no conductors, just fare inspectors, since the entire HSL system, including trams and the metro, has adopted proof-of-payment. (Until the middle of 2017 there was a ticket sales agent in one trainset of each consist, but this was phased out in favor of on-platform ticket machines.) At each boarding door, there are two ticket validators: a contactless one for stored-value fare media and a mechanical one for traditional single-use paper tickets. (HSL does not require passengers with daypasses to validate or “tap in” on trains: the ticket is validated on issue, and passengers can keep it stashed away until asked by a fare inspector to present it.) Under FRA and union work rules, it is likely that at least initially a conductor will be required in each trailing trainset of a consist. For Regional Rail, we will eventually want to attrit the conductors and give those who wish to stay the opportunity to train as an operator or become a fare inspector, which is almost the same job, but there should not be nearly as many fare inspectors needed as conductors today.)
  6. Ingress and egress are efficient. Unlike US systems that operate trains based on conventionally coupled self-propelled coaches, like the LIRR and Metro-North, the FLIRT is a permanently coupled articulated train with no gangways. This also means that there do not need to be doors at the ends of every coach. Door locations and seating layouts can be optimized for increased capacity and faster ingress and egress: you can allow passengers in the articulations (seated and standees, not just passing from one section to another), and the articulations need not be spaced 90 feet apart as those traditional US-designed EMUs do. All of these factors increase the capacity of both the trains and the service, and allow for greatly improved dwell times at busy stations and more consistent running times. Boarding doors on the Sm5 are wide, automatic, and passenger-operated, which speeds alighting, enhances passenger comfort, and saves energy by keeping conditioned air inside the train at stops where few passengers board or alight and not all doors need to be opened.
  7. The electric FLIRT accelerates fast and has a maximum speed of 100 mi/h or more. The maximum speed is actually an option for the customer; European commuter operators typically buy trains that top out at 160 km/h, which is at or above the maximum for most US commuter railroads including most MBTA lines, because the extra horsepower and electrical capacity required are not justified by the increased speed, which is limited more by station spacing than by equipment anyway. However, if you wanted a FLIRT that could pass an Acela Express train on the one 125 mi/h segment of the Northeast Corridor in Massachusetts, they’d be happy to build one and sell it to you. At the other end of the power spectrum, I’ve heard anecdotally that Stadler would likely be interested in making a FLIRT with sufficient battery capacity to perform many low-speed yard moves without the benefit of overhead power, which could reduce implementation costs at least initially — if indeed they aren’t already selling this.
  8. Passenger amenities are plentiful. In addition to an LCD monitor showing route information, train speed, outside temperature, and passenger information at every boarding door, the Sm5 has AC power outlets at nearly every seating position. In the TEXRail diesel variant, both AC outlets and USB power are provided, as well as passenger Wi-Fi. There are small tables at every window seat, perfect for resting a phone or other small electronic device while charging, and the trains themselves have large windows and excellent high-efficiency interior lighting.
  9. The interior layout is bike and stroller-compatible, even at full seating capacity. With the 3+2 seating layout, the corridors are wide, and there is a location at each end, adjacent to the accessible toilet, where bikes and strollers can be hand-held without interfering with passengers walking or wheeling through the train. Ending the off-peak restriction makes active transportation dramatically more viable for many commuters, as they can take the train to work and ride home, or take the train into the city and then use their own bike to get around.
  10. There’s a diesel option. As mentioned above, TEXRail is buying essentially this train, but with a diesel prime mover in the middle. Why is this an advantage, given that the Regional Rail program requires electrification? There are a few reasons. First and foremost, the transition from diesel to electric is going to take some time, and there are lines that may need replacement rolling stock before it’s practical to electrify them — the Needham and Old Colony lines are likely candidates there, with the Needham line eventually being converted to rapid transit and the Old Colony having the single-track bottleneck in Dorchester that will have to be remedied before Regional Rail can be implemented effectively. So it makes sense to have a few diesel-powered trainsets around, particularly with the level of parts and maintenance commonality between the EMU and DMU FLIRT3 systems. Secondly, with some number of diesel vehicles available with a similar capacity to the regular EMUs, it’s much easier to handle situations where the overhead power is unavailable, due to storm, mishap, or ongoing construction. And finally, given a maintenance base that can handle both EMU and DMU systems, it becomes much more practical for MassDOT to operate longer-distance intercity services on lines that will likely never be electrified, like the capeFLYER, East-West Rail to Springfield and Pittsfield, Worcester local commuter service, and commuter service to Nashua and Manchester.

So why not something like a Metro-North M8? Well, point 6 above really gives the show away: there’s no inherent virtue in using a traditional-length rail car if you’re already going to have to significantly upgrade your maintenance facilities to handle EMUs anyway, and eliminating gangways (where passengers are not supposed to sit or loiter) allows greater seating capacity per unit length. It’s not that great a virtue to be able to vary consist length a hundred seats at a time either, and most railroads (like most metros) prefer to leave passenger trains coupled for as long as practical.

Using shorter cars could in theory reduce variable costs, if you were committed to running only 110-seat consists during off-peak hours, but the moment you have a big event (concert, sporting event, convention, political rally) that starts or ends during the off-peak hours, you’ll really appreciate the flexibility offered by running more seats all the time: it means you don’t need to dispatch extra service, with additional operators and their wages, in any but the largest special events. And of course the longer trainsets have only two driver’s positions, at either end of unit, whereas with traditional EMUs you either waste space and lots of money on an operator’s position in every car, or you increase fleet complexity by having to maintain separate “control” and “non-control” cars just as the MBTA does now. (The extra cost is especially problematic with the additional cab equipment required for Positive Train Control: fewer cabs and longer trainsets is better. Compare how the subway cars are managed: there are always two cars “permanently” coupled back-to-back; you can break the pair, but they prefer not to, so that both cars in the pair accumulate wear and need maintenance on the same schedule.)

This is backed up by my simulations of Framingham/Worcester Line loading: in the zone express scenario, the “local” inbound trains that run every 15 minutes from Framingham can make service requirements with only single-trainset consists throughout the day, with peak loading at around 300 passengers (260 seated and 40 standees who got on in Newton and will be getting off in less than 17 minutes) on a couple of inbound rush-hour trains.

So in more general terms, the specification should look something like this:

  • Permanently articulated trainsets between 245 and 267 feet in length. (267 feet is 800 ÷ 3, so the maximum length to fit a three-trainset consist on a standard MBTA high platform.)
  • Seating capacity between 225 and 275 passengers including wheelchair bays, with a 3 + 2 configuration and at least 25 folding single seats, poles and railings for standees.
  • Standard NEC-compatible 25-kilovolt electrification, with PTC/ATC integration in accordance with the MBTA’s existing implementation.
  • At least six wide, automatic release, passenger-operated boarding doors on each side, with 100% barrier free level boarding at all doors and uniform floor height; no traps.
  • Two ADA-compliant toilets.
  • Electronic route and destination signs with the capability to display additional passenger information; automatic stop announcements.
  • Passenger amenities: charging stations in every row of seats, small phone/beverage-cup tables at windows, overhead open racks for bag stowage, at least X contiguous square feet for stroller/bicycle parking (may overlap wheelchair bays); Wi-Fi integration (may be implemented by others depending on how the T is used to procuring this).
  • Vendor should be able to supply a substantially identical configuration with a diesel prime mover (EPA Tier 4 compliant).
  • Minimum acceleration 1 m/s² (for EMUs only); minimum maximum speed 100 mi/h
  • Optional battery pack for off-wire yard moves (trade-off against cost of fully wiring all the yards and layover facilities)

So that’s the “what”, now for the “how”. Obviously, I think this should be put to an RFP as soon as possible, in order to avoid wasteful spending on new obsolete diesel equipment (see the update at the end of Monday’s post on the total bill for Regional Rail). The RFP should specify a delivery date of 2Q2021 for 30 EMUs, with annual options for another 270 cars to be delivered 30 per year starting in 2023, and an option for up to 30 DMUs exercisable between 2021 and 2025. (Why no deliveries in 2022? Because it will take at least a year to fully qualify the new equipment on the Providence Line before the T is ready to accept more trains, and this gives extra time to ramp up platform construction on the Framingham/Worcester line, which would be next after Providence and Fairmount.) The target price per unit for the EMUs should be not more than $10 million. (Helsinki paid under 8 million euros for theirs, adjusting for inflation, and that’s with a nonstandard gauge.)

This entry was posted in Transportation and tagged , , . Bookmark the permalink.

6 Responses to Rolling stock for Regional Rail: What and how to buy

  1. Alon Levy says:

    The main reason not to buy M8s isn’t car length, it’s vehicle mass. The M8s are around 60% heavier than the FLIRTs per unit of train length.

    • That’s mainly because of stainless steel vs. aluminum and composites, right? Or is there some FRA game going on where the feds demanded crazy weight for “safety”?

    • There’s probably a way you could write this into the RFP as an energy efficiency standard. Peak acceleration is exactly tractive effort per unit train mass, so if you set a cap on the energy draw of the whole train, which is easy enough to justify, you also limit the excess weight.

  2. Clem says:

    How do you serve 48” NEC platforms with a low floor FLIRT (615 mm floor height)?

    • You ask them to make it. They didn’t have a broad-gauge FLIRT until the Finns put out their tender in the mid-2000s. Mechanically, you just move some of the train-top equipment like HVAC and power handling to the underside, between trucks. (The MBTA already has experience with this on the subway: Blue Line and Orange Line vehicles are basically the same equipment, but on the lower-floor Blue Line cars, the equipment that sits between trucks on the Orange Line is moved to the roof.)

    • Max Wyss says:

      They already have solved the high level platform issue with the orders for the UK.

Comments are closed.