MBTA "Transformation" (Green Line, Red Line, & Orange Line Transformation Projects)

Correct me if I'm wrong, there is one operator for each trolley car and no plans to change that. Each operator simply needs to go to the other end of the car they are operating. And when type 10s are here, which is what this is planning for, a lot more trains will run as single type 10s or 9/10 combos which the same each operator moving to the end of their car would apply.

The T isn't going to block its only path to One-Person Train Ops on the last line that doesn't yet employ it by kneecapping themselves with a permanent operating kludge. That's cutting off your nose to spite your face. PoP implemented fully lets them get rid of the trailing operator and take advantage of the born OPTO capability that's been built into every car they've ordered since 1986 but never used anywhere except non-revenue moves. They can enormously expand service on a smaller staff by doing that and improve their farebox recovery by a big measure. I don't want to hear one peep about cushy union jobs in the rapid transit div. ever again if we're so hot to de-loop the Central Subway that workplace efficiency is the barter we're permanently giving up. Understand what the forever-and-ever implication is there.


Also...major correction here. . .

Type 10's are going to run as two-car trains most of the time per the GLT horse's mouth as told to the FCMB. Consist sizes are not going to change with the new fleet, making the pursuit of Green Line OPTO ever more a high existential priority. The T is not--and, as we'll see, cannot--make this purchase with a goal of dividing their consist lengths in half. It will be running stretched pairs with double the capacity--2 'supercars' equaling the seating/standee capacity of a 4-car Type 7/8/9 train--as an ironclad prerequisite for all the other big improvements they're doing. Because some of those improvements have other tradeoffs that need to be traded back into balance. The next-gen signal system with auto speed/stop enforcement and computer-primary dispatching won't be able to space trains as closely as today's human-operated single-file. The only way to maintain and ultimately increase Central Subway bandwidth with smidge longer spacing is by aggressively attacking dwell times (and their drag effect on schedules) with more doors, seating, and aisle space per train. This works hand-in-hand with surface PoP and transit priority to solve the "garbage in, garbage out" problem of bunching where branches meet at Kenmore and Copley, so total throughput can be backfilled with more slots at less padding and many fewer canceled trips from late/bunched prior headways aborting the next headway from schedule correction. X feet further spacing on the single-file by not having the human always touching the throttle conversely ends up counterintuitively increasing throughput by being able to slash back schedule padding to the bone. But you need the extra seating capacity to go that hyper-aggressive at the slack padding, so it can't be done by laterally trading over to singlets everywhere but the peakiest-of-peak.

None of it works if train capacity stays near-static to today; the onboard threshold has to get busted through to a whole new level to make this new GLT ops paradigm work. It ends up being a step backwards when the train spacing has to increase and fleet management ends up too target-fixated on minimum-most cars/seating to physically do the job. Supersizing slashes the dwells and (more than) makes up the difference. Therefore, all peak-period trains on all branches will eventually be 2-car 'supertrains' once the backlog of platforms are lengthened. And tame the backlog they must...because this isn't going to work as well as intended if the B and C are malingering for 10 years behind schedule on platform lengthenings.

It also means...yeah, OPTO is also not optional for GLT once they get their PoP game honed enough to leave the rear car's cabin vacant. They need that in order to muscle the human resources around efficiently in this whole new way of doing things.


The only time singlets are going to dominate the service ranks are on the far off-peak when dwells are a total non-factor and headways are usually longer making the train spacing balancing act ^^above^^ moot. Neato...the night shift will cost an insignificant few bucks less to operate. But far off-peak is also when short-turning goes quiet, so running singlets does nothing for speeding up in-car end changes. Everything's running to the ends of branches, little to nothing is turning at loops or crossovers, and they have all the time in the world to do it because headways are throttled-back. Singlet supertrains aren't going to buy us any tangible flex where we need it most. Any time headways are running at full-throttle but loads are not, like the mid-afternoon a couple hours before the peak shift...you're probably still going to see that 2-car service baseline set by the new default train spacing. Short-turn augmentation, on the other hand, is almost entirely a peak-period thing when absolutely everything will be running at 2-car...so we still have to mind our future considerations before getting all demolition-happy with the loops and downtown turnbacks. An end change move time of day when shortie patterns are online is going to be in an OPTO universe where walking outside the train is the only way it can be physically done.

-------------------------

They're putting on such an enormous PR front to explain this to the public because the reasons for supercars require more explanation than just face value. "Divide cars by 2 = SAVINGS!" is what the public's going to assume on-spec. It's not an accurate assumption at all, and not even one-tenth the story of how freaking BIG a sea change GLT is seeking in the operating rules of the Green Line as we've ever known it. The entire service balance is going to be completely different after this, and the supertrains are merely one cog in a whole bunch of other interdependent stuff trading off amongst each other to achieve that new and better service balance. You can't get the full picture of what they're going for by looking at just one cog in a contextual vacuum. If "more singlets = progress" is your first takeaway...STOP...Google for the last GLT public meeting slides...and re-educate. That ain't even the half of it.
 
Last edited:
The next-gen signal system with auto speed/stop enforcement and computer-primary dispatching won't be able to space trains as closely as today's human-operated single-file. The only way to maintain and ultimately increase Central Subway bandwidth with smidge longer spacing is by aggressively attacking dwell times (and their drag effect on schedules) with more doors, seating, and aisle space per train.
I thought a big part of the point of automated dispatching was being able to run closer headways without those silly inattentive humans with their poor reaction time in the mix.
 
I thought a big part of the point of automated dispatching was being able to run closer headways without those silly inattentive humans with their poor reaction time in the mix.

It's complicated.

In most railway applications that's true, because distance-of-separation between trains will loosely correspond to given headway. The Green Line, unfortunately, is a way more chaotic system than most railway applications. It has unusually tiny signal blocks, unusually approximate distance-of-separation between trains, and unusually high number of corrective signal stops to lasso that wildly uneven train spacing in the tunnel into more-or-less even headways at the station platforms. When it succeeds at actually managing the bunching...which it doesn't often-enough succeed at because of "garbage-in/garbage-out" at the portals from poor surface schedule management. Operator rules are also unusually permissive within station blocks, such that they can "run up against the taillights" of the preceding train at the most crowded downtown stations.

When they first studied their next-gen options after the fatal 2007 D Line rear-ender crash got the NTSB on their case about signal auto-enforcement of speeds and stops, the very act of introducing crash protection (any kind: newfangled CBTC, Orange/Red-like simpler ATO, or Blue-like trip stops) direct-conflicted the insanely variable train spacing and tiny signal blocks. Such that, where today the spacing could wax and wane between stations and arrive at the next station rounded up/down to a regular headway...under speed enforcement the in-tunnel spacing had to be more even to be manageable at all. A little less close than closest spacing...but also a more predictable middle-ground than widely-varying near vs. far spacing evening out. Since spacing under today's ops is too inherently chaotic to even out, that meant that the end result was instituting crash protection signaling direct-harmed headways. All of the computer-assisted dispatch optimizations that could help with that technology's capabilities were rendered moot before they ever had a chance because the baseline-most safety features (which...um...we need rather badly!) were what ended up hosing service levels first. Nothing of existing computer signal design had ever tried to tackle a system as ops-unruly as the GL Central Subway, and it showed...painfully...when they evaluated the options.

The T matter-of-fact concluded that they couldn't maintain service levels doing it that way...so back to the drawing board the signal study went, and that was the last we heard about it for a half-dozen plus years. Where that last effort was predicated on default ops--and thus chaotic train spacing--never being able to change because we were locked in by the Central Subway's quirkiness...this new effort starts with a whole new set of assumptions. Mainly, the radical change in operating practices: full surface transit priority, PoP, and stop optimization taming "garbage-in/garbage-out" so Central Subway train spacing is immediately an order of magnitude more predictable. Then you have the huge vehicle-side capacity change aggessively targeting platform dwells in conjunction with the PoP (i.e. the reason why 2x seating ≠ 1/2 the cars per train in real practice). The new signal system will then push the follow-through by eliminating the herky-jerky corrective stops and lengthening some signal blocks that have been kludged way too short. Finally, the computer brains will fine-tune and lend predictive ability to game out the spacing--and uncertainty factors--ahead of the branches meeting at Kenmore Portal in a way today's human dispatchers cannot predict 5 steps ahead of themselves.

In the end you will preserve current headways while having all the necessary safety prevention at long last in-place...which addresses the huge flaw in the headway-killing first study. But while headways will remain par in the sense that trains will arrive on-platform at more or less the same even churn, taming the chaos makes the whole living/breathing system more precise and trackable. Meaning, elimination of nearly all canceled trips due to headway bunching/gapping upsteam netting fewer overcrowded platforms from sudden/inexplicable/excruciating 10 min. service gaps. Which in turn will help the transfer stations greatly by taming their most unpredictable rider swells. Then more ability to do precision-targeted short-turning and varied-up alt routing to address very specific areas of congestion or time-of-day dependent demand pairings...something we'll badly need in order to accommodate expansion like the Urban Ring, and which will move more people total within the Central Subway's headway limit by exponentially enriching the supportable spread of service.

Richer, more dynamic, more precision-targeted...more three-dimensional Green Line service. Within roughly the same 'recommended' headway. Only the 'recommended' headway is going to now correspond to the 'real' headway way, way more often than today's chaos...which works hand-in-hand at clearing away the extra headroom for more precision and service variety accommodating aggressive expansion. It won't be the "anything goes" line anymore. There'll be order and hierarchy in the dispatch room, and a nonspecific feel of a lot more overall order/less seat-of-pants chaos from the passenger's viewpoint.
 
Last edited:
I noticed that the Ball Square Station drawing calls out space for future platform lengthening, but it’s going to be 225 feet on day 1, right? I don’t recall anything in the GLT presentations mentioning 3-car type 10 consists, so why would the platform ever need to be lengthened?
 
I noticed that the Ball Square Station drawing calls out space for future platform lengthening, but it’s going to be 225 feet on day 1, right? I don’t recall anything in the GLT presentations mentioning 3-car type 10 consists, so why would the platform ever need to be lengthened?

Century-level future-proofing for HRT conversion. No conceivable scenario for it right now, but if it's only a matter of spacing tracks wide around an island you reserve all the extra space you can for the hell of it.
 
I noticed that the Ball Square Station drawing calls out space for future platform lengthening, but it’s going to be 225 feet on day 1, right? I don’t recall anything in the GLT presentations mentioning 3-car type 10 consists, so why would the platform ever need to be lengthened?
The GLX design was locked in before the Type 10s, so it was unclear if far future would demand 4 sets of existing cars, which would require 300ft. But now it’s totally irrelevant.
 
As posted here, the Orange Line repair project is starting up again with shutdowns over the next two months.
 
So lurking around, I found this guy who seem to did some back-of-the-napkin math.


If this guy is right, we'll need 5-6 years of weekend downtown shutdowns. Is this what we're looking at? Shutting down whole weekends in the Downtown core to replace ~300'-~500' feet of track against 58,000' of track for the Orange Line alone?

If this guy is right, then the start of the weekend Downtown shutdowns was not a "ripe-the-banndaid" moment to finally start to bring service back to someone respectable. It's the start of a new even lower-service normal. And that amount of track they're replacing, is 300'-500' a number indicating a make-the-most-of-it-round-the-clock replacement work? Or more indicative of a cost-cutting measure of track replacement that could be achieved overnight? Or is it neither, like it is round-the-clock, but incredibly slow (or is actually even fast)?

I don't know, but that's why I post here. Is this what we're looking at?
 
Well, there are a lot of things to consider here, but some really quick math expanding off what he did:
1578672183123.png

Leads to between ~60% (at 400ft/weekend) to ~15% (tripling productivity to 1500ft) of our weekends having the RL shut down between 2020 and 2025.

I expanded it out to weeks in the Crazy Transit Pitches-esque (and coinciding Boston transit riot inducing) event that they shut down lines for a week, in which case it still takes 12 weeks to complete at best case, or 44 at worst case. Though, I suppose you may be able to get a lot more done with a week's worth of time, and the rate can likely be a lot higher.


I'd like to see what their plan is for the "substitute service." that he mentions. If anybody can find the presentation online, can you link it here? Tried looking for it, but found nothing...
 
If this guy is right, then the start of the weekend Downtown shutdowns was not a "ripe-the-banndaid" moment to finally start to bring service back to someone respectable. It's the start of a new even lower-service normal. And that amount of track they're replacing, is 300'-500' a number indicating a make-the-most-of-it-round-the-clock replacement work? Or more indicative of a cost-cutting measure of track replacement that could be achieved overnight? Or is it neither, like it is round-the-clock, but incredibly slow (or is actually even fast)?

I don't know, but that's why I post here. Is this what we're looking at?

This was my concern after looking at what happened with WMATA.

"Rip-the-bandaid" became "normal ops"

My problem is that the MBTA hasn't detailed how they are setting up the work. How many hours are actually getting used?

IE, previously, they had 4 hours of overnight work time available.

A weekend shutdown, with a single 8-hour shift on Saturday and one on Sunday successfullu fits the claim that they are drastically increasing work time. They can also claim to cut costs because they are not paying overnight labor rates.

But that's incredibly ineffective if they have 52 hours to work with and they're only using 16.

What exactly is the MBTA doing?
 
This was my concern after looking at what happened with WMATA.

"Rip-the-bandaid" became "normal ops"

My problem is that the MBTA hasn't detailed how they are setting up the work. How many hours are actually getting used?

IE, previously, they had 4 hours of overnight work time available.

A weekend shutdown, with a single 8-hour shift on Saturday and one on Sunday successfullu fits the claim that they are drastically increasing work time. They can also claim to cut costs because they are not paying overnight labor rates.

But that's incredibly ineffective if they have 52 hours to work with and they're only using 16.

What exactly is the MBTA doing?

I don't remember where but I did read somewhere when these were announced that the whole point was they could use the whole window, that there are certain fixed setup and take down procedures that minimize what can get done overnight and make the longer work blocks more efficient.
 
I don't remember where but I did read somewhere when these were announced that the whole point was they could use the whole window, that there are certain fixed setup and take down procedures that minimize what can get done overnight and make the longer work blocks more efficient.

The set up and take down procedures leave about 2 effective hours of work, and that's if they're 100% on top of everything. The window opens greatly for an entire weekend shutdown as jass pointed out. I'm not familiar with the logistics of weekend shutdowns, but another thing they have to consider on night work is the order in which equipment and material is brought through the system. Once work starts, there's a high chance there's no turning back to Cabot or Wellington to get other equipment/material, as you're likely going to be blocked by another crew/your own crew's machinery. The logistics is another limiter to the amount of work that can be completed. I imagine logistics on weekend-long shutdowns are a different beast, but still a beast to be tackled.
 
The logistics of underground central tunnel work are also much harder to coordinate than outer edges work with more space to move around and bring in and out. So in theory Northern OL work and Southern RL work should go quicker.

However, the red line has a lot more tunnel, and the orange line has everyone's favorite floating slabs in the Southwest Corridor that are always a weekend job regardless of efficiency.
 
I have wondered why the floating slabs can’t be cast elsewhere, trucked in, and completely swapped during the weekend shutdowns. Instead, they appear to be jackhammering and partially rebuilding the slabs in place.
 
I have wondered why the floating slabs can’t be cast elsewhere, trucked in, and completely swapped during the weekend shutdowns. Instead, they appear to be jackhammering and partially rebuilding the slabs in place.

These are questions I hope we can get an answer. If the answers really are like "we're only working 16 a weekend" or "it cost too much to truck in floating slabs and do it Fast-14 style", then we finally (to my knowledge) have something tangible "answer". I have seen plenty of calls for increase funding repairs and maintence - many call for new taxes. But the rebuttals have been things like "MBTA couldn't start enough projects to use up the current maintenance funding" and "we have a problem of sheer backlog rather than funding to repair". And that line of argument, I haven't see a counterpoint. You can't throw more money make the new Orange Line trains deliver faster.

But if they are eliminating faster repairs option like putting in pre-cast full slabs or only working 16 hours a weekend shutdown, then that would be intolerable.

Of course, we don't have the information. It's also possible they are working all 52 hours with full urgency that should be reflect after disasters like the derailment. Then it's very well we would have to live with 5-6 years of half of all weekend with Downtown shutdowns. Then so be it. Can't complain both about everything breaking down while also dealing with the logistics of repairing.

But in irony, I hope it's the former. So we can demand faster approaches. And personally, a genuine example to the what extra funds would allow when people talk about taxes rather than just "congestion tolls would go into a lock box for maintence funding"

In all seriousness, we have groups with lines that can ask these questions now. So are we going to ask it?
 
I get that Orange is narrower (at the floor level) than the Red, but why do the cars need to be *shorter*?

The El had tight curves that required short cars. Where is the limiting curve on the Orange today?

(And in the long run I get that open gangways are the solution to car length, not curve elimination)
 
I get that Orange is narrower (at the floor level) than the Red, but why do the cars need to be *shorter*?

The El had tight curves that required short cars. Where is the limiting curve on the Orange today?

(And in the long run I get that open gangways are the solution to car length, not curve elimination)

Chinatown-Back Bay is the ruling curve. Dudley Sq. loop was the ruling curve when the 01200's were ordered.

Dimensional increase doesn't really rate high. The remaining 1908 downtown stations are really hard to platform-lengthen, and because of the difference in doors the seating capacity per-car is the same as the new Red cars. Unless you can lengthen enough to get a RL-match door configuration (probably too big an ask) without seat reductions, chasing lengthenings down to the last +1 seat really isn't going to return the investment. Besides, there's still remaining headway-improvement slack to chase so dimensional considerations don't rate high until you top out at 3 mins.

Now, Blue is one where retiring Bowdoin Loop and evaluating curve-easing at State in pursuit of Orange dimensions (or as close as humanly possible) does open up another capacity gear to reach for. In part because those cars are so much significantly lower-seating than the others.
 
I have wondered why the floating slabs can’t be cast elsewhere, trucked in, and completely swapped during the weekend shutdowns. Instead, they appear to be jackhammering and partially rebuilding the slabs in place.
At a semi-educated guess, I’d say it’s because in order to completely replace the slab, you have to remove the rails; you can’t slide it out to the side because of the tunnel walls. But you can rebuild in-place without touching the rails except maybe jacking them up a bit.
 
At a semi-educated guess, I’d say it’s because in order to completely replace the slab, you have to remove the rails; you can’t slide it out to the side because of the tunnel walls. But you can rebuild in-place without touching the rails except maybe jacking them up a bit.
That's exactly it. It's continuously welded rail so they jack it up over the individual slab and work one-by-one. Cutting the rail speeds the slab replacements but causes additional complications in having to splice in new ribbon that may make it unable to get a weekend shutdown wrapped in time for Mon. morning.
 

Back
Top