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

Asking a probably stupid question - is there no way to install a switchover between Park and Boylston to allow the train to simply reverse direction at Park? The T only uses 1 of 2 Lechmere-bound platforms as it is (and the one they don't use is currently dedicated to the loop) so having the train wait long enough for the driver to run to the other end seems plausible...

It's generally weird to me that the T would keep any loops, and they're going to an in-reverse-out model at Boston College. Why not here?
Not really. The support beams in the station are really inconvenient for picking ideal spots for a crossover install, it takes too much longer to reverse vs. loop when the operator has to physically change to the opposite-end cabin so ends up reducing capacity vs. the current plan, and the inbound fence side would lose the ability to stack 2 trains in a row like the other side if extra buffer space had to be reserved for a reversing crossover. Future expansion is going to require more fluid ops than that, so it boils down to whether the margins are OK enough to delete the loop entirely or if the throttle is still needed in the face of increasing traffic: either as delay bail-out or as a new TBD breed of short-turn service.

For new TBD breeds of short-turn service, it doesn't even have to be 'expansion'-expansion. Recall that pre-1985 the E turned all Arborway trains at Park but interlined alternating Heath-turning service thru to Lechmere at rush hour. You could see something like that return sooner than later with a relatively minor tweak like E-to-D connecting trackage interlining a Huntington Ave. peak headway-booster out to Brookline Village outbound to Reservoir or bent back in to Kenmore Loop as a circuit. That could change up the inbound-side turnbacks where Park ends up going online for 3 hours each peak to supplement the GLX termini and shorten some routings for extra fluidity during rush. The GLT team will have to consider both the shorter-term use options like that and the much longer-term expansion considerations like Green-Transitway off Boylston, Urban Ring LRT, and a Needham Branch off the D that end up entirely re-drawing where B/C/D/E and the GLX branches each regularly turn.
 
Asking a probably stupid question - is there no way to install a switchover between Park and Boylston to allow the train to simply reverse direction at Park? The T only uses 1 of 2 Lechmere-bound platforms as it is (and the one they don't use is currently dedicated to the loop) so having the train wait long enough for the driver to run to the other end seems plausible...

It's generally weird to me that the T would keep any loops, and they're going to an in-reverse-out model at Boston College. Why not here?
Doesn't the proposed GLT track (depicted below, cloned from upthread) address this issue, even if it has a very different physical solution?
It essentially invites more trains to turn (still on a loop, admittedly) at Government Center
Terminology: Let's use the T's terms "fence track" (for the inner tracks) and "wall track" (for the outer ones)

im]
 
Last edited:
Thanks for the consolidated thread. For the B branch stop rebuild/removal, do we have any updates on construction?
 
It's generally weird to me that the T would keep any loops, and they're going to an in-reverse-out model at Boston College. Why not here?

An in-reverse-out model means moving drivers between cabs and restarting the train. Not an overly long process but certainly not quick. That's fine for terminus stations where there are layovers and often multiple trains waiting to go out but somewhere like Park St I just don't think there is time to block a track for that long to perform that process.
 
Also, it would result in outbound trains leaving from the inbound side of the station.
 
It would make sense to get rid of the Park Street loop with the GLX. A big reason is that rush hour trains entering North Station from Somerville will likely be at capacity, whereas now they're pretty empty but fill up quickly with CR passengers.
 
With new office towers at Haymarket North Station and Lechmere, it may never again pay to turn short at Park Street.

I'd say anything that currently turns at Park Street should turn at government center. Anything that turns the government center should turn it North Station, and anything that turns at North Station you should probably run to Union square in the future.
 
With new office towers at Haymarket North Station and Lechmere, it may never again pay to turn short at Park Street.

I'd say anything that currently turns at Park Street should turn at government center. Anything that turns the government center should turn it North Station, and anything that turns at North Station you should probably run to Union square in the future.
Also NorthPoint is seeing lots of employment growth, it's basically becoming an extension of Kendall.
 
With new office towers at Haymarket North Station and Lechmere, it may never again pay to turn short at Park Street.

I'd say anything that currently turns at Park Street should turn at government center. Anything that turns the government center should turn it North Station, and anything that turns at North Station you should probably run to Union square in the future.

Green Line isn't exactly set up to just keep pushing everything forward-forward-forward so every service pattern runs a marathon. The lost art of short-turning that BERy and the MTA did so well and the MBTA almost from Day 1 ran in pants-shitting terror from is arguably a big reason why the Central Subway is a lot more brittle and less flexible than it was >60 years ago when it juggled 8 surface branches with aplomb. Precision turning is where you mesh routing variety and assign headway-boosters to where they're needed. We not only may not have enough of that, but with the expansion dreams we have it's going to be necessary to get back into the act of hyper-local service supplements and more dynamic pair-matching patterns.

As noted above, stub-end turnbacks are fine at the ends of lines and North Stations where there's a waiting car supply backstopping the somewhat time-consuming end-change, but absent an adjacent yard and in the midst of heavy traffic you need loops to turn fast. Ours unfortunately are on the tight side. So we're in a terrible hurry to tear our Park Loop. Well...what about Brattle Loop which is only a couple feet of radius wider but is most likely going to be needed for rush-hour Medford shorties in a few years because Riverside-Route 16 is too big an ask to sustain at 6 min. headway over that long a distance under peak loading? Are we going to slap artificial restrictions on Brattle for emergency or non-revenue use only and end up running headlong into a problem when Somerville growth goes nuclear?

Or what about ambitous future expansion considerations that the Green Line could handle under GLT traffic management? The Urban Ring is probably going to come with ironclad requirement of turning Cambridge-Lechmere trains @ GC from the north end. What about Kenmore Loop revenue service? If you hook the UR in at BU you're probably going to be running E-to-D circuit service turning at Kenmore to bolster the cross-platform Longwood transfers on all those 66 riders who'll be migrating over to LRT out of Allston, especially since the SW quadrant Kenmore-Dudley Ring will have to be BRT and could use the service augmentation on that first Longwood leg to keep the buses clean for thru-Dudley passengers. Are we going to permit that, or is Kenmore Loop going to be consigned forever to restricted non-revenue use only? Can we answer the question of what patterns might need a Park St. turn in the future before we get too hot to rip it out? Is that where Urban Ring inbound from Kenmore turns on 'circuit' service while inbound from Lechmere boomerang turns at Brattle? Needham and Dudley Sq.: who's the highest priority for a GC-turn slot, but does one have to stand back at Park to do it fluidly? What about Forest Hills redux? If we've made peace with streetcar extension to Hyde Square we're already halfway back to Arborway; does Huntington get fileted Forest Hills-Park and Brookline Village/etc.-Union like the old double-barreled Arborway-Park and Heath-Lechmere peak service of pre-'85? Not all of these are necessarily going to get built, but if you're looking to delete Central Subway infrastructure that's been there since Day 1 in 1897 you better have your 50-year traffic modeling considerations sorted to absolute confidence.


The hard part here is that the 2 ft. difference in radius between seemingly revenue-expendable Park Loop and seemingly revenue not-expendable Brattle Loop isn't a great enough difference to suddenly get us importing a wholly unmodified world best-seller Siemens S70 Avanto like we're bumming off the assembly line for Seattle's order. It's fairly clear that customization is going to have to make up some of the difference no matter what. What GLT hopes is that it can be customization within the off-shelf design rather than cutting the other way: another Boston design-build "Type" series of all low-floor that has to lead with another new truck design based from 'our' system and which merely aims to generify the crap out of supply chain to pound the lifetime costs in-line with industry standards. Either scenario would be a massive improvement over today's sustainability, so by no means does having to for-real "Type 10 it" mean we lost. A Boston-first design is clearly the less satisfying outcome for the GLT task force, but one they'll peacefully accept if just no practical alternative because they think they've hit on a formula for doing it way better than ever before.

It would be preferable if the resulting customization were more like a new "Avanto Tight" flavor that's just a new narrower-turning form factor in the Siemens product catalogue rather than a truck/carbody design that's not organic-Avanto at the root because it had to be led by the nose through redesign by Boston specs. Siemens sells LRV's on three hemispheres of planet Earth. They have legitimately good shot of scoring some follow-on orders from the 'Tight' form factor they just so happen to come up with because Boston's got a kingmaker's 200+ car order up for grabs. It's well worth the R&D of adding a new dimensional flavor to the regular lineup because we happen to be buying in bulk and have some specialer needs to tend to. Somebody on those 3 hemispheres has tight street intersections begging to be marketed to with a flavor that can cut harder corners than the rest but otherwise doesn't have any major design deviations from the rest of the family. The GLT team would much prefer to shop for options "marketed to Boston" from within the family lineup rather than "designed FOR Boston" and fingers-crossed somebody else will like Boston's design.

But no question...if every single loop on the subway other than GC-inbound had to be either ripped up or be slapped with a non-revenue forever restriction for the sake of ordering a generic 'Tight' instead of a "Type...", the constraints on future service are going to be difficult to swallow and will lower the ceiling on feasible expansion. In that scenario taking a hugely improved "Type" car for Boston with heavy self-policing during design to generify the parts supply chain ends up hands-down the better bet. We can't be so target-fixated on off-shelf for off-shelf's sake that it's going to make routing patterns too rigid in the face of growth to accommodate dynamic expansion and a full arsenal of headway-boosting tricks. Something has to give, and if "Type"-ing it with a better blueprint this time still ends up a stratospheric improvement over today...the only bottom line that ultimately matters is what a stratospheric improvement over today it is.
 
Last edited:
Wouldn't ultimately be the best but assuming not doable be to turn all short run B/C/D/E trains at a loop at Haymarket or North Station and all short run Union and Medford Tufts trains at a loop at Park so that each short train hits all Red/Blue/Orange as not to overload downtown stations with 1 and or 2 stop hops to make their connections?

F-Line probably knows exactly why we can't turn after Park going west. I'm assuming Boylston is too close to Park for it. I'm actually surprised there wasn't a loop that way when they first built the Green Line considering all the surface lines were probably used to get to downtown.
 
Could you do an end-change on the straight part of the Brattle loop? Coming up the hill (out the portal in the picture below) in on what is currently the "to Haymarket" side and then back out, using it in the normal direction?

If the Brattle loop needs to be lengthened to accommodate a 2-car Type 10, can it be lengthened on the straight part (toward Haymarket) rather than the curved?

Image result for brattle loop
 

Attachments

  • 1578062457720.jpeg
    1578062457720.jpeg
    8.5 KB · Views: 197
1578148701785.png


What about using the Tremont St tunnel south of Boylston to turn back GLX trains, and eliminate the Brattle Loop that way? Not only would the loop be eliminated, but the Red Line connection is maintained despite short turning downtown.

Credit to Van for the original map which I modified for clarity.
 
Turning in the Tremont St tunnel also has the advantage of putting all 4 tracks into use (Park to Boylston) and using both sides of the platforms at Boylston.
 
View attachment 2391

What about using the Tremont St tunnel south of Boylston to turn back GLX trains, and eliminate the Brattle Loop that way? Not only would the loop be eliminated, but the Red Line connection is maintained despite short turning downtown.

Credit to Van for the original map which I modified for clarity.
50-year considerations: what's going to be built off there crowding out the short-turns?

-- Green-Seaport quite likely. That's enough service to take up 2 of 4 tracks by its lonesome if the Transitway has capacity to pair-match with 2 branch patterns.

-- Washington St. LRT conversion.

-- Relocated E off Copley Jct. If that + the Dudley route equal 2 patterns on one side while 2 patterns matched to the Seaport go to the other side, that's not much left for reversing on-platform.

Granted, the main gain from reopening the Tremont tunnel is that it's capacity for interling 2 northern branches + 2 UR branches. But if short-turning for precision, delay recovery, and targeted rush hour supplements is an important feature to keep in-pocket--as important north-south as it is from the west--do you really want to be retiring loop infrastructure?

Again...not a matter of which projects get built so much as what are the implications to any of them later for decisions made now to delete infrastructure. It's tough. This is why there are outer limits to how far they can go in the name of off-shelf for off-shelf's sake. You can draw up scenarios where ANY of these loops (including Kenmore) end up strategic holds.
 
Could you do an end-change on the straight part of the Brattle loop? Coming up the hill (out the portal in the picture below) in on what is currently the "to Haymarket" side and then back out, using it in the normal direction?

If the Brattle loop needs to be lengthened to accommodate a 2-car Type 10, can it be lengthened on the straight part (toward Haymarket) rather than the curved?

View attachment 2359

No...it wouldn't be possible to do the operator change-ends safely as a regular practice. In-tunnel yards like North Station and all similar stub-ends on the HRT lines have operator platforms next to the stub tracks for walking safely between ends in the middle of an enclosed tunnel. In the 1963-construction GC-Haymarket northbound tunnel it's just the two tracks side-by-side at minimum separation with no physical barrier...not even the wrought iron fence that's there between tracks on the 1897-construction southbound tunnel. Operators would have to exit out the right-handed doors straight into oncoming traffic. And even if you do shiv a fence through there it's extremely narrow passage with risk of bonking head or snagging clothing on things, or getting hit with falling ice from a passing train (which gets deposited all over the subway after accumulating on the roofs during time running on surface). The only time crossovers in an enclosed space are ever used for revenue turnbacks are during service shutdowns (planned or unplanned), where you have an armada of inspectors with flashlights and safety gear stationed in the tunnel escorting the operators around during the end change and punitive bulletin speed restrictions in place for anything that may be moving on the other track.

OSHA going to frown heavily upon trying that as a regular practice anywhere there isn't room for an operator-passage platform. And the union will, for good reason, come in right behind them. But in all likelihood the T's own legal dept. will put the kibbosh on such proposals on liability grounds before it ever makes it far enough along to get leaked to the employees. As for notching the wall for putting in an operator platform...well, once you've invited those costs into the picture all efforts boomerang right over to apportioning the same money to examine notching the wall for loop curve-easing instead, so the idea won't get far enough along to begin with before being immediately redirected to more productive avenues for a fix.
 
Dare I ask why operators need to leave the train to walk to the other end? Couldn't they simply walk through the passenger compartment to the other cab? Am I missing something?
 
Dare I ask why operators need to leave the train to walk to the other end? Couldn't they simply walk through the passenger compartment to the other cab? Am I missing something?

It's not possible to walk between trolley cars. They're fully closed units.
 
It's not possible to walk between trolley cars. They're fully closed units.
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.
 

Back
Top