Fairmount Line Discussion

[The EGE]

Another meeting about BHA on the 15th: http://www.mbta.com/about_the_mbta/publ ... 6442452870

[Arlington]

Proposals are due 12/30/2014 in response to this RFP now posted in the MBTA Business Center

Description: The MBTA seeks to Purchase Multiple High Floor Diesel Units (DMUs) Vehicles in support of the Fairmount Line Service Improvement Project in strict accordance with the MBTA Specifications.

Category: Revenue Vehicles (Bus Heavy Rail...)

Buyer: Aidan Flynn | 617-222-5893 | AFlynn (email)
Documents:
141-14 Public Notice

[ns3010]

Proposals are due 12/30/2014 in response to this RFP now posted in the MBTA Business Center

Description: The MBTA seeks to Purchase Multiple High Floor Diesel Units (DMUs) Vehicles in support of the Fairmount Line Service Improvement Project in strict accordance with the MBTA Specifications.

Category: Revenue Vehicles (Bus Heavy Rail...)

Buyer: Aidan Flynn | 617-222-5893 | AFlynn (email)
Documents:
141-14 Public Notice

Wow, there goes all my previous doubt on whether or not this would actually happen... Let's hope the specifications aren't anything too crazy, and they could just use the same Nippon-Sharyo DMUs that SMART and Metrolinx are getting (although I've heard that initial testing has shown some issues with the hydraulic transmissions)

[F-line to Dudley via Park]

Proposals are due 12/30/2014 in response to this RFP now posted in the MBTA Business Center

Description: The MBTA seeks to Purchase Multiple High Floor Diesel Units (DMUs) Vehicles in support of the Fairmount Line Service Improvement Project in strict accordance with the MBTA Specifications.

Category: Revenue Vehicles (Bus Heavy Rail...)

Buyer: Aidan Flynn | 617-222-5893 | AFlynn (email)
Documents:
141-14 Public Notice

Wow, there goes all my previous doubt on whether or not this would actually happen... Let's hope the specifications aren't anything too crazy, and they could just use the same Nippon-Sharyo DMUs that SMART and Metrolinx are getting (although I've heard that initial testing has shown some issues with the hydraulic transmissions)

No, let's hope they actually follow through on the service plan for Fairmount. If they don't run the line at the frequencies they promised, DMU's are going to be the biggest waste of money ever. The vehicles don't show their advantages until they get at or near a real clock-facing schedule, not the pathetic Fairmount headways they're years late in bolstering. They need to show some commitment on the ops side. I don't care if that means stretching push-pull to the schedule density limit before they're allowed to cut the checks to the DMU manufacturer. I have almost zero confidence that the follow-through on the schedule is going to be real. And if they're saying they can't increase the schedule to what it said in the implementation plan until they have DMU's on the property...they're lying. I'd rather see them prove it...then buy the DMU's...rather than deal with the probability that these things are going to be some political white elephant.


And frankly, until they figure out how they're going to fund all their conventional equipment replacements--all ~200 remaining single-level coaches and ~45 remaining F40/Geep locos--that are on a collision course in the Fleet Plan for a 2020 retirement date...they shouldn't be buying anything extracurricular. That's an enormous amount of equipment to stay on top of and not push their luck in extended life. If DMU's in any way, shape, or form raid sparse funds that otherwise would've gone to those commitments, this is going to end up being a net negative for the system. Another reason I'd rather wait to see them prove they're good to their word on the schedule with existing equipment before risking a funding raid on shiny things-for-shiny things'-sake.

[deathtopumpkins]

But F-line, that would make sense. This is the T we're talking about here.

[ns3010]

I don't disagree that the DMU's are wasteful and unnecessary, that's why I was hoping all along that this would never actually happen. I've said it before, and I'll say it again, but they could easily run three car (even two if they wanted to) PP sets, with no real safety risk. It is done every day by other major and smaller roads, such as NJT and M-N, with no issues at all.

I'd rather see them spend whatever absurd amount of money they're about to spend on this on new PP coaches instead. Like you said, the single levels aren't going to last forever, and most of the cars are at or approaching retirement age. Had they spent this money on new PPs instead, they could use some of the remaining Pullmans and Bombs to run the high-frequency Fairmount shuttles. While some of the aging F40 stretches and Geeps aren't in the best condition and can't handle the 7-8 car bilevel sets as well as the HSPs, they would still be fine handling 2-3-4 car Fairmount scoots.

[Arlington]

I think the move to DMUs is political, but that politics aren't always bad. DMUs just send a more "urban" and "transity" vibe, which may be worth experimenting with. Either they'll work and the T will buy more, or they won't and the T can probably find someone to take them off our hands.

[F-line to Dudley via Park]

I don't disagree that the DMU's are wasteful and unnecessary, that's why I was hoping all along that this would never actually happen. I've said it before, and I'll say it again, but they could easily run three car (even two if they wanted to) PP sets, with no real safety risk. It is done every day by other major and smaller roads, such as NJT and M-N, with no issues at all.

I'd rather see them spend whatever absurd amount of money they're about to spend on this on new PP coaches instead. Like you said, the single levels aren't going to last forever, and most of the cars are at or approaching retirement age. Had they spent this money on new PPs instead, they could use some of the remaining Pullmans and Bombs to run the high-frequency Fairmount shuttles. While some of the aging F40 stretches and Geeps aren't in the best condition and can't handle the 7-8 car bilevel sets as well as the HSPs, they would still be fine handling 2-3-4 car Fairmount scoots.

It's chicken-and-egg. Once you do get 15 minute clock-facing frequencies enough to encourage critical-mass usage, the interior configuration of a push-pull does suck. Vestibule doors are lousy for quick boarding, 3 x 2 seating is awful for routes that have constant stop-to-stop churnover, and bi-levels (once the last singles are purged) are cumbersome for the same reasons. At that level of service you do need a more rapid-transit like interior configuration to make the passenger flow work optimally.

THAT SAID. . .

None of that matters until the schedules are 100% of goal. 10 out of 10 people will take the push-pull that runs on-time and at plannable intervals over the specialty vehicle that doesn't run nearly as often as they promised.

None of that matters until the zone fares are redistributed equitably across all lines in the Indigo network, which they are not on the 2024 Indigo map.

None of that matters until the commuter rail is Charlie'd and these Indigo routes can be self-contained on a single combo card. So many people using Fairmount and the other proposed Indigo routes transfer to buses that it's pointless if they have to pay the onboard ticket premium or whip out a smartphone.

It *probably* doesn't matter until there's some transfer equitability such as tap-on/tap-off timed transfers to rapid transit at South Station, Back Bay, and possibly even Yawkey. Otherwise the utilization is going to be a bit below goal.

The T has a bad reputation on following through with promises and needs to deliver on all of the above before buying shiny things. And the riding public has a low tolerance for BS and isn't likely to be distracted for very long by shiny things.

Push-pulls can cover the schedule buildout to 100% as a bridge fleet. It is only when the riders start coming that the interior layout starts showing its limitations. It will take a few years to build the ridership to that critical mass where you need the DMU interior layout...so why does this purchase need to be front-loaded before 2020 at all? It can come after they've determined budget apportionment for the next round of push-pull fleet replacement. The rush should be raising eyebrows.

Ops optimization plays a greater role in achieving the full implementation plan than vehicles. So if the CR equipment is maintained so inflexibly that they can't run anything less than 4 cars, and they don't break consists appropriately by time of day, and they overstaff trains with more conductors than necessary...what exactly about a new vehicle purchase changes said inflexibility and inefficiency? Can we get Keolis and the T to substantiate how they're going to be more nimble before giving them new toys? Because new toys don't make a lick of difference if they can't/won't be nimble.

Why the hell are they doing a RFP before the FRA announces revised specs on crashworthiness...currently scheduled for 2015 but probably being delivered a couple years later? They'll have wider selection to choose from putting this off 2 years, and the RFP paperwork may have to get done over all the same if the revised regs really are as significant as hoped. For what PR snow job are they in such a damn hurry for?

Why are they doing a RFP before SMART and Metrolinx get their vehicles on the road. Those two services already have questionable reputations for over-customization, something the T needs desperately to start avoiding given that it doesn't have any feet left to shoot holes in over vehicle overcustomization disasters. Can we see some of these FRA-compliants--the new ones, not the Colorado Railcar dead-enders--performing in actual revenue service first?



I don't know. Their rep is so tarnished I just put the odds very low that this is going to be the panacea they're over-hyping it to be. It reminds me too much of the BRT bubble 15 years ago and what a letdown that half-finished effort ended up...unicorn vehicles and all. If the service isn't there, the service isn't there. If the fares and transfer options are broken, people won't be fooled. If the budget's blown they can't keep up with the rest of their necessary fleet replacements...systemwide service is going to suffer and that'll be its own discouragement to ridership whether or not Fairmount itself or the DMU's themselves have good uptime, because everything painted purple-tinged is going to share in that bad reputation. I can't shake that baked-in cynicism the way this is unfolding. All-flash, no substance. And the substance really has nothing to do with whether DMU's are good vehicles and everything to do with frequency and reliability of service. Which makes the state's hyper-focus on the vehicles being THE service a colossal--and inherently dishonest--distraction.

[Bramdeisroberts]

Exactly, and here's my hunch.

Everyone here is complaining that the T is pursuing DMU's while ignoring the push-pulls and the impending need for a proper replacement for the F40's/Geeps and the flats. What nobody here seems to want to talk about is the possibility that the DMU's ARE the replacement for the older push-pull hardware. Is it really that far outside the realm of possibility that the T could be looking to drastically downsize the push-pull fleet, saving it for high-density work (think HSP46's pulling all-bilevel sets) where needed, while replacing everything else with DMU-based service?

The precedent is there if you look at the other "premier" commuter rail systems, even in North America. Metronorth prioritizes MU service, using push-pulls where MU's don't make sense, and the same goes for the LIRR and NJT, where push-pull service is almost exclusively high-density bilevel based trains, with MU's operating on a "lower capacity, higher-frequency" basis. SEPTA is all MU-based and is able to pull the same ridership as the T over a smaller area served (and one with far less of a traffic nightmare than we currently have). Sure, the other systems use EMU's instead of DMU's, but for the T, DMU's offer a great chance at proving the MU-based service model here without having the insane up-front costs of electrification.

Who knows what the plan actually is, but this all is a pretty intriguing possibility to me, and I, for one, am all for the T sticking their necks out and having even an ounce of real vision rather than watching them continue to pull the "this is the way we have always done it and the way we've always done it is the way we always will" routine that they've been so stuck in for the last quarter-century.

[F-line to Dudley via Park]

Exactly, and here's my hunch.

Everyone here is complaining that the T is pursuing DMU's while ignoring the push-pulls and the impending need for a proper replacement for the F40's/Geeps and the flats. What nobody here seems to want to talk about is the possibility that the DMU's ARE the replacement for the older push-pull hardware. Is it really that far outside the realm of possibility that the T could be looking to drastically downsize the push-pull fleet, saving it for high-density work (think HSP46's pulling all-bilevel sets) where needed, while replacing everything else with DMU-based service?

The precedent is there if you look at the other "premier" commuter rail systems, even in North America. Metronorth prioritizes MU service, using push-pulls where MU's don't make sense, and the same goes for the LIRR and NJT, where push-pull service is almost exclusively high-density bilevel based trains, with MU's operating on a "lower capacity, higher-frequency" basis. SEPTA is all MU-based and is able to pull the same ridership as the T over a smaller area served (and one with far less of a traffic nightmare than we currently have). Sure, the other systems use EMU's instead of DMU's, but for the T, DMU's offer a great chance at proving the MU-based service model here without having the insane up-front costs of electrification.

Who knows what the plan actually is, but this all is a pretty intriguing possibility to me, and I, for one, am all for the T sticking their necks out and having even an ounce of real vision rather than watching them continue to pull the "this is the way we have always done it and the way we've always done it is the way we always will" routine that they've been so stuck in for the last quarter-century.

DMU's of the type they are buying here will never replace conventional commuter rail equipment. Ever.

There is no way DMU's can keep up with the capacity demands being put on conventional commuter rail, to the point where they could feasibly displace those vehicles. This isn't 1955 and B&M going to a 100% RDC fleet in an era where ridership is collapsing and lines are easiest to run with single cars, 2-4 cars lashed up, and very limited schedules.

-- Those Worcester and Providence peak trains need 7 bi-levels pulled by a 4600 HP beast. Right now, today.
-- Plenty of other routes run 6 cars and badly need the bi-level infusion on their most overcrowded runs.
-- Major future growth is coming rather soon for all manner of Fitchburg schedules, off-peak Worcesters, Middleborough after the Buzzards Bay extension opens and with downtown Brockton finally getting some major tall-building mixed-use development rising.
-- Per-train Lowell ridership is likely to explode to an extreme if the Nashua extension happens. The NH Main has the highest untapped growth potential after Worcester, and it isn't overly expensive to go out and grab it. Per-train Haverhill ridership is likely to grow a healthy bit when Bradford gets replaced by a bigger layover supporting fatter schedules (possibly augmented by more Anderson expresses), and Plaistow would add a lot of riders. If Foxboro happens the Franklin main ridership in Walpole, Norwood, and Dedham is going to go through the roof from having 16 round trips double to 32 and making those towns major car-free commute destinations, plus blowing the lid off the 128 park-and-ride at Dedham Corporate. Gains moreso at the existing stops than Foxboro itself lathers on top. And so on, and so on.

How exactly is a single-level DMU with 2 x 2 rapid transit seating and seats removed for extra doors supposed to keep up with the sardine-packing that is an issue on today's ever-increasing capacity push-pulls, let alone the growth that's upcoming. Are we supposed to lengthen every platform on the system to 10-12 cars to keep up with it? How many billions will that cost?

Push-pulls aren't going anywhere. These DMU's are for an ADDED layer of 128-inbound centric service and displacing the very meager Fairmount equipment pool which immediately gets sucked right back up by Worcester and Fitchburg increases. Not 1 single conventional coach is going to be displaced by a DMU here. The RDC era isn't coming back, because the RDC era coincided with when schedule frequency and peak-hour passenger loads were a tiny fraction of what they are today. Every single commuter railroad in North America is junking single-level push-pull cars on an aggressive schedule to go with all bi-level rosters, NJT and Metro North the latest to announce 100% purges over the next 5-10 years. None of them are going to be considering DMU's for anything but specialized service where the specialized nature of the cars fits the service plan. It is bunk overhype that these things are about to take over the world like some sort of universal-solvent rolling stock. They aren't. EMU's...maybe, if new electrification starts piling up by the hundreds of miles in the next 20 years. But electrification is a totally different dynamic and more dramatic change in buying habits than conventional vs. self-propelled diesel.


And...capacity-capacity-capacity is trickling up to the EMU ranks too. NJT's Arrow replacement order is going to be crammed into the MLV coach dimensions. That is part of their 100% bi-levels in 5-8 years strategy. SEPTA could easily tag along and order identical makes for its Silverliner VI order. Caltrain, GO, AMT...they're looking at bi-level EMU's for their officially-proposed or officially-desired electrifications system-wide or system-majority. Metro North and LIRR are quite likely going to be the last users of single-level EMU's when all current generations of self-propelled electrics are overturned. East Side Access doesn't allow anything 1 inch taller than an M7 to fit through the small 1970's-construction part of that tunnel, Metro North piggybacks onto the LIRR car order for its Hudson/Harlem vehicle scale, and the New Haven Line's unique power requirements stuffs too much electronics underneath the M8 carbody to fit the same underneath an MLV carbody. The MTA roads will be the only conventional commuter rail outlier left still running single-levels, because they don't have a choice.

Yes...if the T ever looks at EMU's over push-pulls for Providence and a pool fleet with RIDOT for Providence-Westerly, it's going to be bi-levels probably not much different from the new NJT design (artful Photoshop depiction here: download/file.php?id=5525). Maybe to our taller K-car and Rotem dimensions instead of the shrunken Penn-fitting MLV's that lack headroom and overhead luggage racks...but what they consider would be 95% the same design. And tailored to those crowd-swallower Providence mega-consists, not faux- rapid transit seating where you have to run 12+ cars to fit the same number of bodies. DMU's today--at least the configurations they would want to order for Indigo service--are specialty vehicles for a certain narrow range of optimized, high-frequency, short-distance, high stop-to-stop ridership overturn. Where you better damn well have the frequencies, equitable fares, and service optimization to draw the riders...otherwise the vehicles don't make any difference whatsoever. DMU's are not universal commuter rail vehicles, and they are wholly inadequate for serving the capacity needs of trains to/from the 'burbs currently run with push-pulls of >4 cars (or even 4 bi-levels with few unfilled seats). The only agencies seriously biting on 100% DMU fleets are the upstarts like SMART and WES that have no other vehicle scale and are at least 1 future generation of vehicles away from ever needing monster per-consist capacity. Those carriers won't have to ponder single- vs. bi-level, packing-bodies vs. easy on/off for another 18-20 years when it's time to wholesale-replace their entire 1st-generation fleets. The T and other Top 10 ridership carriers have no such luxury to cleanroom a fleet from ground zero when a majority of equipment already has to take crowds well beyond an Indigo-configuration DMU's capabilities. It's a big mistake to conflate this purchase with the universal RDC fleet from a drastically different era in transit history. Today's commuter rail is drastically, drastically different from when the last self-powered RDC made a revenue run way back in 1982.

[Bramdeisroberts]

I realize that Worcester, Providence, Nashua, Bradford (or if we're talking 30-year pipe dreams, an eastern route reactivation to Portsmouth) will always generate the ridership to fill big bilevel consists.

That said, what's to stop the T from purchasing 2 or 3-car trainsets and running them as 6-car sets at increased frequencies during peak hours on the 2nd-tier lines like the Eastern Route, Fitchburg, or Franklin, then splitting them up for off-peak service. Take the Fitchburg line, where the 6-flat sets as ~50-minute headways are approaching sardine-can conditions during rush hour. You could keep running push-pulls at current frequencies and add cars/switch to bilevels, or you could switch to running 6-car DMU sets (2 3-car sets or 3 2-car sets) which would get you 75% of the capacity of a 6-flat consist, but run them at 150% of the frequency. That would land you a net gain on the lines that can handle the extra trains, so you'd be saving your push-pull sets for the lines with either the ridership to justify big push-pulls or where bottlenecks prevent higher frequencies. Hell, of you ordered Silverliner-style carbodies, the low platforms wouldn't be the end of the world either.

During off-peak, split the big DMU sets (this is a BIG if, though I feel like it's much less of a case of "nobody does it" than it is a case of "the T doesn't do it") and now you have extra hardware to keep frequencies up off-peak while keeping the really big push-pull sets out of service off-peak.

It's all hypothetical, and there are some big logistical hurdles to bridge, but none of it would require treading over anything approaching new ground for any of the world's top-flight regional and commuter systems, though given that this is the T we're talking about, even that might be a tall order...

[F-line to Dudley via Park]

I realize that Worcester, Providence, Nashua, Bradford (or if we're talking 30-year pipe dreams, an eastern route reactivation to Portsmouth) will always generate the ridership to fill big bilevel consists.

That said, what's to stop the T from purchasing 2 or 3-car trainsets and running them as 6-car sets at increased frequencies during peak hours on the 2nd-tier lines like the Eastern Route, Fitchburg, or Franklin, then splitting them up for off-peak service. Take the Fitchburg line, where the 6-flat sets as ~50-minute headways are approaching sardine-can conditions during rush hour. You could keep running push-pulls at current frequencies and add cars/switch to bilevels, or you could switch to running 6-car DMU sets (2 3-car sets or 3 2-car sets) which would get you 75% of the capacity of a 6-flat consist, but run them at 150% of the frequency. That would land you a net gain on the lines that can handle the extra trains, so you'd be saving your push-pull sets for the lines with either the ridership to justify big push-pulls or where bottlenecks prevent higher frequencies. Hell, of you ordered Silverliner-style carbodies, the low platforms wouldn't be the end of the world either.

Because that's not the way fleet scale works at large commuter railroads. Look at how NJT, Metro North, LIRR, and even Amtrak have structured their fleet plans. They all play up a move to fleet commonality, simplifying their rosters around common makes, increasing seating capacity to run more efficiently configured consists, and to value increased ops efficiency over fleet diversity. And are very explicit about the economics of standardization being way better than fleet fragmentation, which is why they are ponying up the big bucks right now to overturn their fleets.

-- NJT is purging 100% of its single-level coaches for MLV's and 100% of its Arrows for MLV EMU's, all with the same seating configuration. They're standardizing on single-source manufacturers (Bombardier) for all of their coaches and ALP-xx locos, and could very well do the same with this EMU purchase. Their current motley assortment of equipment will by decade's end be either an MLV coach or an MLV EMU (both pretty much identical on the interior), with locomotive roster simplified down to just the ALP-45/46's and PL42's. Even on the Princeton Dinky...a bi-level EMU is going to be the ride seating overkill be damned because the only way the Dink continues to exist is if it can cheaply share off the systemwide equipment pool. They aren't going to keep 2-4 Arrow singlets on the roster in rotation just for the sake of that tiny thing.

-- MNRR and LIRR are standardizing their bi-level coach and dual-mode loco makes, and for the first time unifying all of their non-EMU equipment in common orders for both railroads. Single-source EMU's now (Kawasaki) with lots of parts commonality on the M7/M9 and M8 designs. Even down to ordering only third rail shoes that can work on either railroad so they can stock only 1 warehouse full of the same part. And now going to a common MLV-dimension coach and common dual-mode locomotive. Probably with the coaches being off-shelf NJT/MARC/AMT-style MLV's from Bombardier or a copycat instead of in-house design like the C3's (which are all being junked).

-- Amtrak went with one supplier for next-gen locos: Siemens. Sprinters replace all electric push-pulls, and they've got the Charger (which shares some Sprinter components) on-order for the state-sponsored services. If/when the national option orders expand out the Chargers replace all P42/P40's, F59PHI's, and Dash 8 revenue fleets for one make. Probably with a dual-mode version soon to follow to replace the P32's. All non-East Coast cars push-pull cars are being standardized on various configurations of the Superliner. All of the Heritage fleet is going Viewliner. When the Amfleets and Horizons are finally up for replacement, the entire single-level fleet is almost certainly going to Viewliner-based coaches. They are very explicit about fleet commonality being the goal for everything that doesn't involve specialty trainsets (Acela/next-gen HSR fitted to NEC quirks, passive-tilt Talgos for higher speeds on particularly curvy corridors, etc.). Specialty trainsets have to have very specific goals to execute on, and extra due diligence put behind making sure those purchases fit their intended service patterns with maximal efficiency.



That is one hell of a large precedent to be setting here with the fleet plans of such massively large carriers. It is a BIG burden-of-proof to overcome for cutting against that now-established grain. If there is such overwhelming evidence for fleet standardization improving operating economics, you have to have one hell of an airtight argument for bucking that trend and introducing more niche vehicles into the fleets of a Top 10-size operator. Even the T is following the standardization logic with its bi-level purchases around a common design and loco purchases (if the HSP-46 platform has enough legs for follow-on orders). If the T is going to introduce specialty DMU's, they have to show their metrics on how it's going to improve the service plan above-and-beyond generic vehicles. And the only way it's justifiable against this mountain of precedent is if the service plan executes spot-on enough to exploit that specialty niche.

The economic argument is O-V-E-R about mixing fleets elsewhere. It may look superficially like a possible good idea to run low-capacity DMU's off-peak system-wide, but the real data doesn't back that up at all and argues the exact opposite effect. Absolutely, run more efficiently put-together push-pulls on the off-peak to push-pull destinations, but you don't get ahead by prying rapid transit-optimized cars outside their territorial range when the outer layover yards aren't set up for it, crew qualifications are different, fueling cycles are different, and the seating configuration is wasted on the different audience. As I said, this is not the RDC era. This is the commuter rail mode's equivalent of the SL1 airport buses on the bus mode...narrow-targeted service (on the Silver Line: dual-modes with luggage racks; on the commuter rail: extra doors and rapid-transit seating configurations). You don't run Silver Line Airport buses on the #1 down Mass Ave. any more than you would run Fairmount DMU's to Forge Park. The specialty features are totally wasted outside their narrow target corridors, and the operating economics get all murky and eventually really start to suck.


As is there are only a few narrowly-targeted factors favoring an Indigo DMU configuration over a push-pull configuration for Fairmount, etc. 15-minute clock-facing headways with high stop-to-stop ridership turnover...and the type of transit riders who would utilize that service...need:
-- no more than 2 x 2 seating. 3 x 2 commuter rail seating is optimized for people staying put from boarding to the terminal or terminal to exit, not getting up frequently from intermediate stop to intermediate stop.
-- extra center doors and open layout with elimination of the vestibules. Likewise to facilitate easier onboard movement when headed to/from the doors on quick trips from intermediate stop to intermediate stop.
-- no narrow bi-level stairs, for the same onboard foot traffic reason as above.

If the service levels do not exploit this layout, it's wasted:
-- If people have to plan trips around the paper schedule ("Oh, I've got to leave now if I'm going to catch Train #xxx!") instead of a reliably recurring clock-facing headway, they won't ride the train for short-hop trips between intermediates. They'll take the clock-facing bus instead for quick trips, and only opt for commuter rail the same way the outside-128'ers do: to/from home stop and the terminal.
-- If there is not fare equity or transfer options, people won't ride the train for short-hop trips between intermediates. They'll take the clock-facing bus for quick trips and only ride to/from the terminal like conventional commuter rail. Same as above.
-- If the headways aren't frequent enough...same as above. The audience rides it like conventional commuter rail, not quasi- rapid transit.

In any of these scenarios where the service levels fall short, the riders stop doing quick trips and plan it all around the terminal not unlike the suburbanites out in Attleboro do, stay seated longer, and don't have to jockey for position at the doors much. Then it makes zero difference if you're running a short push-pull consist on the Fairmount instead. All of the special features are wasted. THAT'S how narrow a service target they have to hit to make the vehicles a worthwhile purchase. They are not every-vehicles that can be plugged in anywhere. They are "Indigo Line" vehicles that only pay off when run on the Indigo Line service plan. Otherwise they would not be buying them at all.

During off-peak, split the big DMU sets (this is a BIG if, though I feel like it's much less of a case of "nobody does it" than it is a case of "the T doesn't do it") and now you have extra hardware to keep frequencies up off-peak while keeping the really big push-pull sets out of service off-peak.

It's all hypothetical, and there are some big logistical hurdles to bridge, but none of it would require treading over anything approaching new ground for any of the world's top-flight regional and commuter systems, though given that this is the T we're talking about, even that might be a tall order...

And that's a little dangerous logic, equating "nobody does it" with innovation. This is not un-studied territory for the big boys. The sheer breadth of detail in the fleet strategies of the big carriers--yes, including the T's overall push-pull procurement strategy--says the Top 10 passenger roads have a very good idea internally where the efficiencies are. And that bucking the trend requires unusually slam-dunk numbers to justify it. Substantiation Massachusetts pols and T officials are chronically unable to provide. So no...this is a "beware the outlier" situation where marching to a different drummer does not make one look smart. It makes one look stupid when it doesn't work. See the Silver Line lessons for that.

One one of the reasons why none of the big boys have adopted DMU's on their conventional (i.e. non- RiverLINE) systems is because they aren't every-vehicles on systems that are overwhelmingly dominated by every-vehicles. The applications where they are being given serious consideration tend to be of the narrow-target Indigo/inside-128 variety: short shuttles ranging close to the city, near/mid-distance reverse commute, and transfer-optimized off-peak services (e.g. fattened outer LIRR off-peak dinky schedules coordinated with MUCH fattened electric-territory transfer trains to exploit ESA). And none have bitten the bullet yet because it's tough to find enough of those routes to hang a fleet purchase on. For example, I don't think a LIRR Oyster Bay DMU dinky and a Greenport Scoot DMU dinky--desireable in-concept as they may sound--are going to cut it for the MTA unless they can find twice that scale elsewhere on the LIRR/MNRR systems to hang a new fleet purchase on. Then consider that not every system has potential routes that would lend themselves to the same type of vehicle configuration...you don't buy the same seat and door configuration for a dense-stop route in the urban core mimicking a rapid transit line as you do for a dinky on the outskirts of the system to transfer inbound. You'd have to buy two different types, which harms the economics. The MTA roads and NJT probably aren't going to do it because not enough of their candidate lines have the same characteristics as the T's various Indigo candidates. So not only does the T have to stick to the Indigo script faithfully, but it has to follow through on the other inside-128 Indigo routes to give this fleet some lasting scale...and not start getting ideas of mixing their audiences too much.


And even then...not one single piece of push-pull equipment is going to get displaced by this purchase because Indigo done right is such a wholly other thing to conventional commuter rail and 100% of the past-128 demand will still be riding--and needing--conventional capacity in increasing numbers. If anything displaces a chunk of the push-pull fleet it's going to be EMU every-vehicles...since EMU's (especially the bi-level ones about to hit the East Coast) are much better suited to being configured as every-vehicles with their performance advantage in all situations.

[Bramdeisroberts]

Wouldn't a far simpler reason why DMU's haven't been implemented in the post-RDC world be because most of the top-tier commuter agencies with the ridership to justify the higher frequencies that DMU's allow all inherited ex-NYC/PRR electrification infrastructure, making EMU's the no-brainer option? That's certainly the case with the MNCR/LIRR/NJT/SEPTA, and the same goes for the Illinois/Northern Indiana lines.

What's left in North America that honestly has similar per-mile ridership #'s to the T? There's Caltrain, which can and does fill its all-bilevel sets to crush capacity to the point that they're now electrifying. There's AMT, which electrified. Then there's the diesel METRA lines and GO transit, which both are closer in terms of ridership to Caltrain than they are to the T and have much more pressing platform constraints that severely limit service frequency and force them to increase capacity instead of frequency to add service.

I think that the DMU in the US was a real victim of bad luck and unfortunate circumstances. The RDC was a great, if flawed product, and the SPV-2000 was a turkey at the worst possible time, where all the railroads that had the need for and could afford MU's were all under wires, and the roads that could have used them (like the T) simply didn't have the funding to take a gamble on a new product with any chance that it'd be a lemon.

Once Budd went belly-up, they took with them all of that experience building FRA-compliant DMU's and thanks to those regulations, it only made sense for the Bombardier/Adtranz/Alstom/Kawasakis of the world to do that R&D legwork if there was a big enough order to justify it. Since the commuter railroads with the money were all running under wires or over third rails, or in the GO Transit/Caltrain situation, there never was that big buyer that could snag one of the big manufacturers for a proper DMU build, and the smaller startup commuter railroads just didn't have the money/numbers to afford new push-pull equipment let alone clean-sheet DMU's from someone like BBD or Siemens.

Meanwhile the MBTA/CR is a bit of an odd duck because while there certainly has been a need for MU-like service frequency in places like Fairmount, Chelsea/Lynn, and Waltham, we never had electrified service there like they do in NYC, Philly, or DC. Electrification a la AMT or Caltrain is the obvious answer, but there's never been the political will to spend the money on increased frequency, much less on any new hardware or infrastructure to serve our fairly unique needs. And so, to me, it's no surprise that with all of that in mind, we have yet to see DMU's properly implemented anywhere, much less in our backyard.

[The EGE]

It's worth noting as well that DMUs only have any advantage at all over push-pulls on closely packed stops. On the southside that's only Fairmount (maybe to 128 or Dedham or Dedham Corp Center) and Worcester (Needham should be rapid transit by the time DMUs are systemwide); all four northside lines inside 128 are good candidates if you throw in infills*. That's, at most, six route stretching no further than 128 park-and-rides.

Over average stop spacing longer than say two to four miles, loco-hauled is vastly more efficient per seat mile. One big powerplant versus many smaller ones. And you only have to maintain that powerplant.


*Weston/128, Clematis Brook, Alewife, Union Square on the Fitchburg; Montvale Ave (Woburn) and GLX transfer (probably Lowell Street) on the Lowell; Revere and South Salem on the Eastern Route

[F-line to Dudley via Park]

It's worth noting as well that DMUs only have any advantage at all over push-pulls on closely packed stops. On the southside that's only Fairmount (maybe to 128 or Dedham or Dedham Corp Center) and Worcester (Needham should be rapid transit by the time DMUs are systemwide); all four northside lines inside 128 are good candidates if you throw in infills*. That's, at most, six route stretching no further than 128 park-and-rides.

Over average stop spacing longer than say two to four miles, loco-hauled is vastly more efficient per seat mile. One big powerplant versus many smaller ones. And you only have to maintain that powerplant.


*Weston/128, Clematis Brook, Alewife, Union Square on the Fitchburg; Montvale Ave (Woburn) and GLX transfer (probably Lowell Street) on the Lowell; Revere and South Salem on the Eastern Route

And the acceleration difference really is overrated. I think perception is greater than reality because of how long we've had to put up with those gimp F40's straining to pull consists way longer, way taller/heavier, and way more packed full with human flesh than they were originally purchased to haul. The difference in acceleration between an HSP-46 pulling anything vs. an F40/Geep pulling that same anything is greater than the difference between an overpowered HSP-46 pulling a minimum-length Fairmount consist vs. a Fairmount DMU. The DMU's savings pulling out from a stop all fall within the range of standard-issue schedule padding for variable dwell times. Have a person in a wheelchair on the platform need assistance getting on that one time, or hold the doors open for that last person sprinting up the ramp to get on that one time...that's about equivalent to the time difference you're talking in acceleration savings. That difference lives wholly inside the built-in schedule margin for error. It's a total non-factor in whether or not to purchase DMU's. It really is only that optimized interior and door layout once you hit service levels where the generic coaches start showing their limitations. If it were only about horses and stops/starts on a dime, a long-term commitment to a fleet of 4600 HP minimum push-pull power and smarter selection of consist lengths (maybe lifting the sub- 4-car restriction at long last) accomplishes same schedule performance within the default margin for error and 100% of the headways. That's why 'real' Indigo service levels meriting a dedicated fleet are such a narrow target the T has to prove they can execute 100% before we entrust them with play money for vehicles.


EMU's...yeah, they're different. The acceleration difference above-and-beyond all else does save time on short runs, as well as on those 8-car Providence behemoths when the self-propelled trainset can rev up to 90 MPH from a dead stop worlds faster than anything else.

Wouldn't a far simpler reason why DMU's haven't been implemented in the post-RDC world be because most of the top-tier commuter agencies with the ridership to justify the higher frequencies that DMU's allow all inherited ex-NYC/PRR electrification infrastructure, making EMU's the no-brainer option? That's certainly the case with the MNCR/LIRR/NJT/SEPTA, and the same goes for the Illinois/Northern Indiana lines.

No. Because there's some plenty high-frequency push-pull routes out there on NJT, Metra, and GO. Especially in places where an inner zone has one mainline serving multiple branches where your de facto diesel-belcher frequencies in the equivalent "inside-128" space are a no-fooling 15 minutes. Electrification--beyond the earliest short-lived experiments--never stretched a whole lot beyond where it currently goes. And between the Harlem Line's 1984 White Plains-Southeast extension, the New Haven-Boston electrification, and some minor post-1980's NJT infill the electric network has gained more route miles than it lost from whatever old PRR electric branchlines fell by the wayside between the Depression and the RR bankruptcies.

Check the all-time RDC roster: http://en.wikipedia.org/wiki/Budd_Rail_ ... nal_owners. They were primarily a Northeastern phenomenon other than a fairly significant Canadian, Brazilian, and Cuban (!) investment in them. And Canada mainly ran them in long-distance service because of the number of very remote regions that had gov't-mandated service, so CN's and CP's usage profile for DMU's was quite different (though they did see commuter service in present-day GO and AMT territory). 35% of all the RDC's that were ever built were on the B&M roster; they were 'the' major outlier railroad at going whole-hog with them for near-100% of the fleet. What Boston had from the 50's into the T-logoed late-70's was very atypical of commuter ops everywhere else. No one else went so heavily to DMU's for such a long time. Other than here, they saw heaviest usage clustered around the Mid-Atlantic...NYC to DC on roads like NYNH&H, NY Central, and B&O. A lot of co-mingling with electric territory. They were very sparsely deployed anywhere else.

Pretty much everywhere but this region: steam gave way directly to diesel push-pull and self-propelled units were never part of the equation.

What's left in North America that honestly has similar per-mile ridership #'s to the T? There's Caltrain, which can and does fill its all-bilevel sets to crush capacity to the point that they're now electrifying. There's AMT, which electrified. Then there's the diesel METRA lines and GO transit, which both are closer in terms of ridership to Caltrain than they are to the T and have much more pressing platform constraints that severely limit service frequency and force them to increase capacity instead of frequency to add service.

85% of the daily commuter rail riders in North America ride LIRR (#1), NJT (#2), MNRR (#3), Metra (#4), GO (#5), SEPTA (#6), and the T (#7). Yes, Metra and GO (which, yes, is on a frequency-increase binge in addition to planning electrification) both blow the T out of the water on total riders; Metra has double the daily riders, GO with over 1.5 times as many. Ridership-per-mile is a flawed metric when some systems (Metra) cover way more total miles because Greater Chicago is way more spread out than Greater Boston, and some systems (Caltrain) pack all their service patterns onto 1 line instead of 13. Extend the list out to the full Top 10 (#8 Ferrocarril Suburbano, #9 AMT, #10 Caltrain) and you're at over 92% of total North American ridership. Throw in Metrolink (#11) and MARC (#12) to include every agency that does 30,000+ daily riders (it's a cavernous drop-off after MARC to 15K or less) and you're over 96%.

That's every EMU-running railroad on the continent except the South Shore Line (whose ridership really could be lumped in with Metra Electric's), and not a single DMU in sight. The first DMU user on the list, now that Tri-Rail has sold off its Colorado Railcar lemons and gone all push-pull, is TRE at #19 (8000 daily riders).

OK? DMU's serve a very tiny fraction of bottom 2% of ridership on the continent. That is the very definition of niche specialty service. It backs exactly up the big boys' fleet management plans and supporting data: only invest in specialty vehicles when the service is guaranteed to hit its target mark with the specialty vehicles and won't hit that mark running "every-vehicles". Fleet standardization with "every-vehicles" in all other situations where common equipment scale trumps all else on value. DMU's are not a vehicular revolution about to be uncapped and sweep across the continent if only someone were outside-the-box thinking enough. It's been number-crunched to death, and those fleet plans philosophically universal to the top carriers are a correct assessment of cost/benefits. Counterarguments need to show their math against such a mountain of evidence. There are real specialty services out there to be tapped, and Indigo is definitely one of them if the T hits every mark in the service implentation plan. But it is absolutely false that wholesale turnover is going to come to the 98% of the continent's ridership who take an "every-vehicle" for their commuter train. That trending doesn't exist to the degree some people assume it does.

I think that the DMU in the US was a real victim of bad luck and unfortunate circumstances. The RDC was a great, if flawed product, and the SPV-2000 was a turkey at the worst possible time, where all the railroads that had the need for and could afford MU's were all under wires, and the roads that could have used them (like the T) simply didn't have the funding to take a gamble on a new product with any chance that it'd be a lemon.

Once Budd went belly-up, they took with them all of that experience building FRA-compliant DMU's and thanks to those regulations, it only made sense for the Bombardier/Adtranz/Alstom/Kawasakis of the world to do that R&D legwork if there was a big enough order to justify it. Since the commuter railroads with the money were all running under wires or over third rails, or in the GO Transit/Caltrain situation, there never was that big buyer that could snag one of the big manufacturers for a proper DMU build, and the smaller startup commuter railroads just didn't have the money/numbers to afford new push-pull equipment let alone clean-sheet DMU's from someone like BBD or Siemens.

The numbers do not support this. See above. RDC dominance was a uniquely Northeastern thing and ESPECIALLY a Boston/B&M thing. We were the outlier, not the trend-setter. There's no historical precedent for what you're suggesting could've happened had history broken differently.

Meanwhile the MBTA/CR is a bit of an odd duck because while there certainly has been a need for MU-like service frequency in places like Fairmount, Chelsea/Lynn, and Waltham, we never had electrified service there like they do in NYC, Philly, or DC. Electrification a la AMT or Caltrain is the obvious answer, but there's never been the political will to spend the money on increased frequency, much less on any new hardware or infrastructure to serve our fairly unique needs. And so, to me, it's no surprise that with all of that in mind, we have yet to see DMU's properly implemented anywhere, much less in our backyard.

Yes. But I outlined why that was the case. It takes hitting a specific service niche to make them pay off. Otherwise there's no compelling reason to deviate from the scale of a universal "every-vehicle" fleet. The T has that opportunity with the Indigo network clustered around very common usage on all the candidate inside-128 lines.

Other railroads can't find enough commonality. You'd be hard-pressed to match up enough equal-characteristic LIRR, MNRR, or NJT diesel routes that would support that kind of vehicle order. The ones in innermost-shuttle territory would need a more Fairmount-like seat/door configuration, the ones in outermost-shuttle territory (like LIRR/MNRR transfers into electric territory) wouldn't work so well in that same configuration with longer stop spacing, longer sitting, more luggage storage needs, etc. So they either buy two different interior configurations of DMU and lose too much vehicle scale, hedge on one configuration that's going to be good for one user but kinda stinky for another...or, just keep using the every-vehicles because they can't economically split the difference with specialty cars.

And other-other railroads that do have that commonality tend to be clustered in that 2% at the bottom where they're starting real small.


I'm sorry...all of this stuff has been thought of long before. A commuter railroad of any heft is going to butter its bread with every-vehicles...be they push-pull or EMU's, but probably similarly configured bi-levels of either type if they're not in a one-of-a-kind clearance or AC/DC power situation like the MTA with the M# cars. And specialty purchases are going to break along the consistent guidelines in these fleet-management plans: buy niche vehicles only when the niche vehicles hit paydirt that the every-vehicles can't. If every large commuter railroad and Amtrak are staking their futures to this philosophy and being very careful about that paydirt part viz-a-viz specialty rolling stock, counter-examples kind of need to prove their numbers.

The good news is a fully-realized Indigo service plan proves those numbers. We are one of the only installations in the country where the stars align enough to carve out a significant-scale DMU network with the same specialty configuration serving that many potential routes. That's luck of geography more than trend-setting. What they have not--at all, at the most basic level--proved is that they have the means to implement that fully-realized service plan. And if they don't, taking the plunge is financial and operational suicide. It doesn't have to be that way. They can substantiate a lot more detail about HOW the service plan is going to be realized...at those clock-facing frequencies...with that necessary fare flexibility. And then start walking the talk with push-pulls. And then...and only then...buying the specialty DMU's when the service plan has some present-day reality to it and is not such a risky proposition of will-they/won't-they commit on the ops side. That is the only way they're going to avoid shooting themselves in the exact way those big RRs' fleet strategies warn against shooting yourself.

As I said a few posts up, it is gonna take a few years to build up the ridership to the point where this new mode is attracting short-trip riders at the Talbot Ave.'s of the world. The crowds aren't going to be there at critical mass for 3-5 years if they started a 15-minute Fairmount schedule with Charlie fixins' tomorrow morning. There is no reason in the world why they have to wait until new vehicles are on the property to get started. Make the decision in 2020 when service is established enough that the optimized interior DOES offer up a real extra gear for escalating ridership. That moment is not going to happen in the 2010's decade even if they start delivering on long-deferred promises. Get the bloody schedule running like it should be tomorrow and worry about utter vehicular perfection when the momentum catches up. Doing it the other way around is rank political dishonesty on their part that is probably going to cost the T more riders than it gains if it raids the equipment replacement budget to the point where old junk starts breaking down en masse...again.