Green Line Reconfiguration

[Riverside]

I think the main point here that I didn't realize beforehand is: You're implicitly assuming some Huntington trains would regularly turn at Park St inner loop under Criss-Cross. Now with this detail, I'm fully onboard, and I do agree that Criss-Cross can still be a capacity boost even for Huntington.

Ah oops, I see. In fact, I’m actually suggesting that Criss Cross would either loop some trains at Park, or would add a northbound crossover north of the station, or both. (This detail is tucked subtly into the track diagrams above, but I should have mentioned it more clearly. )

Previously, my mental image was that Criss-Cross would leave the Park St loop unused (except for run-as-directed trips), but Tripld uses it to terminate B and C trains. In that scenario, we're essentially comparing M=40 for Tripod to M=30 for Criss-Cross, and it does become a notable negative for Huntington.

Yeah. And it’s interesting to note that I think this is actually how it’s being used today — the Inner Loop is just being used for RADs. I dunno what to say about that.

So there's still a notable difference between LRT and HRT for an apples-to-apples comparison.

Thanks for finding those numbers! Yes, I think this is a fair point. For the moment, my thinking remains that the best path to high frequencies is a simple network topology, but obviously my thinking on that has evolved over the years, so perhaps I’ll be convinced to go the other way again!

I don't see what you meant by "keeping a version of the E in place"? Under Criss-Cross, the bottleneck isn't anywhere close to Huntington tracks itself: it's on the Tremont St subway, either at Park St or Government Center.

Sorry, this was a thought I had toward the end of writing the novel (lol) so I didn’t explain it very clearly. Using the Criss Cross designations, I’m suggesting:

• Quad tracks between Symphony and Back Bay/Copley: one set (Green) goes to Park St, the other (Gold) to the Seaport
• At Symphony, Green tracks continue underground in a new subway
• Gold tracks rise to surface through existing (or slightly reworked) portal at Northeastern, and run in the existing (or slightly reworked) reservation at least until Brigham Circle
• The “Fenway Branch”, which on my diagrams above is used to serve Heath St + Hyde Sq, would instead loop at Kenmore (and potentially be truncated at Brookline Village as a shuttle, or something, with the Gold Line continuing at surface level to Heath — that’s what I meant by “keep today’s E Line intact”)
• Fenway Branch Gold Line trains would therefore be swapped out for Huntington Surface trains
• Provides a 1SR between Huntington and the Seaport, and provides additional capacity along Huntington itself

Alternatively, the same topology could be used with a full-length quad track Huntington Subway.

maybe? It wouldn't need to dive under Boylston station though so maybe not, I'm not sure. I'm also not sure if the plan was designed with the Washington St Subway in mind or not.

More to come later whenever I have time to read a novel lol.

Oh that’s a fair point. I went back and looked at the diagrams, and sure enough they suggest crossing just south of the station. You need to underpin the tunnels to the Pleasant Street portal, but that’s probably easier.

That being said, I think if you did that, it would be much harder to use all four tracks at Boylston, which is hardly fatal, but still a drawback.

 

[TheRatmeister]

Oh that’s a fair point. I went back and looked at the diagrams, and sure enough they suggest crossing just south of the station. You need to underpin the tunnels to the Pleasant Street portal, but that’s probably easier.

That being said, I think if you did that, it would be much harder to use all four tracks at Boylston, which is hardly fatal, but still a drawback.

Here's the plate, for the Post Office Sq extension:

And it looks a lot like Silver Line Phase 3. Which then makes me wonder: Has the difficulty of that potentially been overstated by this board? Did the cost overruns and challenges come more from mismanagement than actual technical difficulty?

Assuming it is possible, what about a service plan like this?

This requires a flying junction at Boylston to make it work but it would allivate the loop-congestion problems that tripod has while giving more people one seat rides, or at least very easy two seat rides. So that leaves two unsolved engineering questions:

1. Can another flying junction be added to Boylston?
2. Can an extension along Essex St navigate the Big Dig tunnels while (presumably) staying above the Washington St subway? (Presumably yes if that was the plan for SL phase 3.)

Another alternative would be to send the Tremont St Subway under Charles St, connecting to a Stuart St Subway, which would briefly be quad-tracked as part of a flying junction. From there, Seaport bound trains would continue along... somehow. This seems like the best case for a surface connector, where a portal on Surface Rd around Tufts St would hug the west side of Surface Road to avoid dealing with vehicle crossings, then on the north side of Kneeland St just immediately dive back underground as soon as the big dig tunnels are out of the way. That would probably require cutting off the connection between Tyler and Hudson Streets and Kneeland St on the North Side, but if you can do that there should be no vehicle crossings at least. A one lane 'wrap around' connection to Tyler St would be to at least allow through-vehicle access to avoid pain for deliveries, all of which would use Beach St which would be converted to two-way traffic by removing street parking.

 

[Teban54]

Sorry, this was a thought I had toward the end of writing the novel (lol) so I didn’t explain it very clearly. Using the Criss Cross designations, I’m suggesting:

• Quad tracks between Symphony and Back Bay/Copley: one set (Green) goes to Park St, the other (Gold) to the Seaport
• At Symphony, Green tracks continue underground in a new subway
• Gold tracks rise to surface through existing (or slightly reworked) portal at Northeastern, and run in the existing (or slightly reworked) reservation at least until Brigham Circle
• The “Fenway Branch”, which on my diagrams above is used to serve Heath St + Hyde Sq, would instead loop at Kenmore (and potentially be truncated at Brookline Village as a shuttle, or something, with the Gold Line continuing at surface level to Heath — that’s what I meant by “keep today’s E Line intact”)
• Fenway Branch Gold Line trains would therefore be swapped out for Huntington Surface trains
• Provides a 1SR between Huntington and the Seaport, and provides additional capacity along Huntington itself

Alternatively, the same topology could be used with a full-length quad track Huntington Subway.

Something similar to this has been mentioned in my 2x2 plan (posted again here for convenience):

Specifically, the N bullet (Magenta Line) basically does what you said -- except that I have N trains use Huntington Ave subway until Copley, then build flyovers for it to converge to the Kenmore-Seaport trunk (Silver in my diagram).

I don't think there's an inherent need for quad-tracking Symphony-BBY, if the goal is solely to run the Magenta route -- 4 tracks are nice but far from necessary. The capacity pinch will always be at Park-GC, not Symphony-BBY, as long as your Nubian-Tremont trains (my Gold Line) + Huntington-GC trains that don't short-turn at Park) >= Huntington-Seaport trains. Quad-tracking Symphony-BBY helps with reliability (and possibly speed if you're willing to build an express/local setup), but even 2 tracks here will always run under capacity.

A separate question is maintaining a surface connection from Huntington to Heath St, but that's not at odds with 2 Symphony-BBY tracks: you can have the "underground to BV" and "surface to Heath" tracks branch out anywhere west of Symphony.

 

[JeffDowntown]

Here's the plate, for the Post Office Sq extension:
View attachment 60841
And it looks a lot like Silver Line Phase 3. Which then makes me wonder: Has the difficulty of that potentially been overstated by this board? Did the cost overruns and challenges come more from mismanagement than actual technical difficulty?

Assuming it is possible, what about a service plan like this?
View attachment 60840
This requires a flying junction at Boylston to make it work but it would allivate the loop-congestion problems that tripod has while giving more people one seat rides, or at least very easy two seat rides. So that leaves two unsolved engineering questions:

1. Can another flying junction be added to Boylston?
2. Can an extension along Essex St navigate the Big Dig tunnels while (presumably) staying above the Washington St subway? (Presumably yes if that was the plan for SL phase 3.)

Another alternative would be to send the Tremont St Subway under Charles St, connecting to a Stuart St Subway, which would briefly be quad-tracked as part of a flying junction. From there, Seaport bound trains would continue along... somehow. This seems like the best case for a surface connector, where a portal on Surface Rd around Tufts St would hug the west side of Surface Road to avoid dealing with vehicle crossings, then on the north side of Kneeland St just immediately dive back underground as soon as the big dig tunnels are out of the way. That would probably require cutting off the connection between Tyler and Hudson Streets and Kneeland St on the North Side, but if you can do that there should be no vehicle crossings at least. A one lane 'wrap around' connection to Tyler St would be to at least allow through-vehicle access to avoid pain for deliveries, all of which would use Beach St which would be converted to two-way traffic by removing street parking.

Silver Line Phase III was going to be routed under both Boylston and Chinatown stations.

 

[TheRatmeister]

Silver Line Phase III was going to be routed under both Boylston and Chinatown stations.

That was the plan for the Post Office Sq extension too. You can see the elevation chart at the bottom, it would pass below Boylston and Chinatown, but above the RL tunnels. Which is why I'm asking if maybe the difficulty of SL Phase 3 has been overstated, if the plan was to do something very similar in 1913.

 

[F-Line to Dudley]

Silver Line Phase III was going to be routed under both Boylston and Chinatown stations.

...and under the GL tunnel from Boylston Station to Charles St. South.

Here's a render of what Chinatown Under station would've looked like:

...and Boylston Under Loop (old render...from when the South End tunnel was going to underpin the abandoned Tremont trolley tunnel):

EDIT: BTW, here's the SL Phase III 2005 EIS, in-full: https://books.google.com/books?id=-ho3AQAAMAAJ

Some of the proposed grades would top out at 5-5.6%, higher than the 5% recommended grade in the MBTA Silver Line specs guide and only slightly less than the 6% maximum allowable grade. Chinatown's placement was modified a couple times in the scoping to fit within these requirements.

 

[Teban54]

In response to @TheRatmeister's comment, I'll start here:

Assuming it is possible, what about a service plan like this?
View attachment 60840

The spirit of this plan seems identical to my 2x2 plan, in that both aim to enable all 4 connections between pairs of destinations: (Kenmore / Huntington) -- (Park St / Seaport). @Riverside also implied that while he didn't explicitly draw out the 2x2, the Criss-Cross plan should allow easy modifications to add Kenmore-Park and Huntington-Seaport tracks.

So I think there's a broad consensus on a route design level; the remaining discussions mostly concern engineering of 2x2, especially with Kenmore-Seaport.


Essex St in 2025 vs. 1925

Here's the plate, for the Post Office Sq extension:
View attachment 60841
And it looks a lot like Silver Line Phase 3. Which then makes me wonder: Has the difficulty of that potentially been overstated by this board? Did the cost overruns and challenges come more from mismanagement than actual technical difficulty?

I had these thoughts before, but my answer was no: Essex St is probably still very difficult to build, even without mismanagement.

Most challenges that SL Phase 3 faced were with the geography and topology, which remain regardless of planners:

• Steep, near-maximum grade going downhill from South Station all the way to Chinatown and Boylston
• Essex St being very narrow with tall buildings
• Underpinning Boylston and Chinatown, both of which are structurally complex themselves

To make things worse, it's not hard to think that some of these factors have worsened from the era of the Post Office Sq branch being planned. Presumably more buildings have come to existence on Essex St; we also have stricter ADA requirements, etc.


That said... The idea of a Stuart-Kneeland St Subway is actually intriguing, and I had considered similar lines before. I'll write more about this when I have time.

 

[Stlin]

I would generally agree that Essex should be looked at again as a GL option. A few months ago I was looking at the 2005 EIS; What struck me from it was that the "core segment" from Boylston to South was well settled, engineered, and appeared buildable. However, there was considerable controversy about the portal alignment. Keep in mind that this was Phase 3 of the SL - it was the central segment meant to connect the Phase 1 Washington St surface running to the Phase 2 seaport tunnel - it was absolutely implicit and required that SL buses had to have a place to enter the tunnel. Look at the proposed alignments - that Tremont St routing, would have required a bypass tunnel to enable buses to "double back" along the core segment alignment in order to reach the initial portal location on Washington, built in the literal middle of a hospital, then NEMC.

The Columbus Ave portal might be straightforward for a tunnel, but would have added significant surface routing complexity to the SL route. The core segment, including its turn back loop, was 3000ft. Each of these portal routings would have added another 2000ft, and with their curves? No bus was going to be able to navigate it with any sort of speed or efficiency. It would also have encountered considerable underground complexity resulting from crossing the surface under a hospital. But if you can tie into the existing GL tunnel / service using the PO square branch stubs between Arlington & Boylston?

 

[JeffDowntown]

I would generally agree that Essex should be looked at again as a GL option. A few months ago I was looking at the 2005 EIS; What struck me from it was that the "core segment" from Boylston to South was well settled, engineered, and appeared buildable. However, there was considerable controversy about the portal alignment. Keep in mind that this was Phase 3 of the SL - it was the central segment meant to connect the Phase 1 Washington St surface running to the Phase 2 seaport tunnel - it was absolutely implicit and required that SL buses had to have a place to enter the tunnel. Look at the proposed alignments - that Tremont St routing, would have required a bypass tunnel to enable buses to "double back" along the core segment alignment in order to reach the initial portal location on Washington, built in the literal middle of a hospital, then NEMC.
View attachment 60842
The Columbus Ave portal might be straightforward for a tunnel, but would have added significant surface routing complexity to the SL route. The core segment, including its turn back loop, was 3000ft. Each of these portal routings would have added another 2000ft, and with their curves? No bus was going to be able to navigate it with any sort of speed or efficiency. It would also have encountered considerable underground complexity resulting from crossing the surface under a hospital. But if you can tie into the existing GL tunnel / service using the PO square branch stubs between Arlington & Boylston?

I have always wondered if you could rise up into the Green Line tunnel via the location of the former Public Garden Portal, near Arlington Street. The tunnel is wide there with the former portal in the center.

 

[F-Line to Dudley]

I would generally agree that Essex should be looked at again as a GL option. A few months ago I was looking at the 2005 EIS; What struck me from it was that the "core segment" from Boylston to South was well settled, engineered, and appeared buildable. However, there was considerable controversy about the portal alignment. Keep in mind that this was Phase 3 of the SL - it was the central segment meant to connect the Phase 1 Washington St surface running to the Phase 2 seaport tunnel - it was absolutely implicit and required that SL buses had to have a place to enter the tunnel. Look at the proposed alignments - that Tremont St routing, would have required a bypass tunnel to enable buses to "double back" along the core segment alignment in order to reach the initial portal location on Washington, built in the literal middle of a hospital, then NEMC.
View attachment 60842
The Columbus Ave portal might be straightforward for a tunnel, but would have added significant surface routing complexity to the SL route. The core segment, including its turn back loop, was 3000ft. Each of these portal routings would have added another 2000ft, and with their curves? No bus was going to be able to navigate it with any sort of speed or efficiency. It would also have encountered considerable underground complexity resulting from crossing the surface under a hospital. But if you can tie into the existing GL tunnel / service using the PO square branch stubs between Arlington & Boylston?

The grades at the portals were absolute crap. 5.6% for the Tremont portal, 6% for Columbus (I'm not even sure why that one is so steep given the straight runup space). They were totally flailing by this point, and that's where the project started slowly coming unglued to its ultimate demise. The reasoning for the 5% "recommended" grade was that the traction motors in the buses would've had to overpower at >5%, shortening component lifespans on a captive fleet that had to make those grades every single run. Up to 6% was tolerable if T maintenance could keep up with the extra component wear-and-tear...while above that you'd simply be shorting out the motors from the power surge. It wouldn't be any better for BEB's because the traction motors are identical and they'd also have to overcompensate for the added weight of the battery bulk.

I have always wondered if you could rise up into the Green Line tunnel via the location of the former Public Garden Portal, near Arlington Street. The tunnel is wide there with the former portal in the center.

The junction would be functionally pretty elegant, but that's still an awful lot of structural underpinning to blow out costs on. While the core segment was largely settled business, the underpins of Green and Orange + the Essex buildings had already extracted their pound of flesh on the overall budget.

 

[TheRatmeister]

I have always wondered if you could rise up into the Green Line tunnel via the location of the former Public Garden Portal, near Arlington Street. The tunnel is wide there with the former portal in the center.

I'm sure it's possible, some of that space would likely have been used for a Post Office Sq extension.

Steep, near-maximum grade going downhill from South Station all the way to Chinatown and Boylston

It would undeniably be steep but planners in 1913 seemed to think it was doable with the streetcars of the day, and it seems to have been backed up with the SL Phase 3 plans. I don't immediately see a reason not to trust that conclusion. It would probably add cost but so would a longer alignment along Stuart St, so it's pick your poison really.

Essex St being very narrow with tall buildings

Is this really an issue a big issue if it's done with a TBM? Chinatown Station is not exactly short on entrances to connect up with underground, so I'm not sure you need a massive expansion of surface-level station facilities. Again, it wouldn't be cheap but the alternatives aren't cheap either, plus they have pretty lousy connections to the Orange Line at Chinatown or Tufts.

Underpinning Boylston and Chinatown, both of which are structurally complex themselves

Could this be dealt with by just avoiding it? What if the platforms are located between Boylston and Chinatown, connecting both stations? The tunnels obviously still need to get under Chinatown station but that's likely to be significantly easier than a large station cavity.

The core segment, including its turn back loop, was 3000ft. Each of these portal routings would have added another 2000ft, and with their curves? No bus was going to be able to navigate it with any sort of speed or efficiency. It would also have encountered considerable underground complexity resulting from crossing the surface.

The grades at the portals were absolute crap. 5.6% for the Tremont portal, 6% for Columbus (I'm not even sure why that one is so steep given the straight runup space). They were totally flailing by this point, and that's where the project started slowly coming unglued to its ultimate demise

To me at least this says to me that most of the challenges of Phase 3 were the link to Washington St, and the correspondingly complex Boylston station. If you can just, not do that, suddenly it looks way more doable.

 

[F-Line to Dudley]

Is this really an issue a big issue if it's done with a TBM? Chinatown Station is not exactly short on entrances to connect up with underground, so I'm not sure you need a massive expansion of surface-level station facilities. Again, it wouldn't be cheap but the alternatives aren't cheap either, plus they have pretty lousy connections to the Orange Line at Chinatown or Tufts.

Green Line has the same general max allowable grade as Silver Line: 6.0% for up to 2500 ft. length. Recommended max grade is 4.0% for unlimited length, which is actually 1% less than it is for Silver. While LRV's have much more tractive effort than an electric bus, steel wheel slip becomes an issue at particularly steep grades so it ends up more or less a wash. So, no, you can't just TBM at unlimited depth and climb like hell to meet the preset interfaces like the Transitway connection. It has to fit within the max grades, and that's where SL Phase III already tied itself in knots trying to sort out.

Could this be dealt with by just avoiding it? What if the platforms are located between Boylston and Chinatown, connecting both stations? The tunnels obviously still need to get under Chinatown station but that's likely to be significantly easier than a large station cavity.

They already evaluated Plans A thru Z for the Chinatown platform siting on the initial Phase III scoping. It landed where it did with that awkward offset stacked tunnel because that's the only configuration that would fit within the 5% recommended grade. I doubt you're going to try to steepen it for LRV's because it's still a performance hit. And between-block would mean you're adding complexity with the long walkways (and added transfer penalty) to each station stacked onto the same narrow footprint. Given how many iterations it took to even settle the Phase III core alignment, I have to think this was considered at one point and rejected as infeasible.

To me at least this says to me that most of the challenges of Phase 3 were the link to Washington St, and the correspondingly complex Boylston station. If you can just, not do that, suddenly it looks way more doable.

The portals were the sticking point that ultimately killed the project when the FTA withdrew its "recommended" rating, but a whole lot of damage was previously racked up to the budget by trying to fit the core alignment. Yes, you could recycle it for LRT on the grounds that it's "proven engineering-feasible", but would it still be worth it at the going price and with underpinning work yet to come to tie it into the Public Gardens junction? Not exactly clear-cut.

 

[TheRatmeister]

It landed where it did with that awkward offset stacked tunnel because that's the only configuration that would fit within the 5% recommended grade. I doubt you're going to try to steepen it for LRV's because it's still a performance hit. And between-block would mean you're adding complexity with the long walkways (and added transfer penalty) to each station stacked onto the same narrow footprint.

I don't see how moving the Chinatown platforms west would reduce the amount of space available for an incline, if anything it would have the opposite effect, but given the limited space available under the station it would probably just stay the same. (That is to say, if you can fit new platforms under the existing Chinatown, you can probably fit station-less tunnels too.)

So, no, you can't just TBM at unlimited depth and climb like hell to meet the preset interfaces like the Transitway connection. It has to fit within the max grades, and that's where SL Phase III already tied itself in knots trying to sort out.

If you have all of Essex St between Washington St and Kingston St to make that climb it's not a problem. What I'm not sure of is is what the depth of the CAT under Surface Road is, and if the line can fit over it. Even if you need to make the whole climb up from ~40-50ft depth to ~20ft(?) depth between Surface Road and Atlantic Ave, that's still not even as steep as the climb from North Station to Science Park, which, while slow, is within the capabilities of existing and future GL rolling stock.

Given how many iterations it took to even settle the Phase III core alignment, I have to think this was considered at one point and rejected as infeasible.

Maybe, but without seeing the list of station sites considered I don't want to jump to that conclusion immediately. I don't think it really wouldn't have made sense with the loop, so that could have just put it out of the running from the start. Or some big-wig demanding two stops could have also done it, or maybe the shorter platforms for buses just made it the better option.

And between-block would mean you're adding complexity with the long walkways (and added transfer penalty) to each station stacked onto the same narrow footprint.

There is definitely a transfer penalty, but I don't think it's that bad. The block is about 500 ft long, let's target a platform length of 300ft. That leaves roughly a 100ft walkway to Chinatown or Boylston Stations on either side. The complexity of navigating the inbound/outbound platforms at Boylston would probably make that transfer longer, but most GL riders would have other, better options such as just transferring at Arlington or Bay Village. Nubian riders would have a longer transfer but even in the worst case scenario that's only going to be as bad as the longest OL/BL connection at State. Definitely not ideal but also not catastrophically bad.

 

[F-Line to Dudley]

Maybe, but without seeing the list of station sites considered I don't want to jump to that conclusion immediately. I don't think it really wouldn't have made sense with the loop, so that could have just put it out of the running from the start. Or some big-wig demanding two stops could have also done it, or maybe the shorter platforms for buses just made it the better option.

In 2003 a combo station at the mid-block was explicitly considered, as were Alts. with the combo stop and no short-turn loop. All were rejected due to incompatible grades, excessive dwell times, and the fact that it provided zero cost savings over doing 2 stations because of the complexity of the connecting concourses. The 2005 EIS covers this in Section 2.3.4.3, "Screening of Mid-block and No-loop Alternatives". Mid-block would've required a 6.4% grade in violation of the maximum tolerances. The egresses would've been more constrained by the narrower and more densely-abutted width of the street at the mid-block location making it less able to cope with the dwells of the increased ridership from pooling the two stations, and rush-hour platform dwells were expected to exceed the minimum operating headway vs. the 2-station option that spread the transfers and split the dwells. Lengthening the platforms to fit more berthing buses to counteract the headway interference only created punitive bunching that sidelined any conjoined attempts to get rid of the short-turn loop and run all service thru. It got eliminated from any further consideration on all those demerits.

Given that an LRT appendage would likely generate even more ridership than the performance-compromised BRT in a similarly constrained space, it's quite likely that the dwell issue is going to hamper any revived attempt at a combo station. And the grades are what they are, with or without the loop.

 

[Stlin]

I don't see how moving the Chinatown platforms west would reduce the amount of space available for an incline, if anything it would have the opposite effect, but given the limited space available under the station it would probably just stay the same. (That is to say, if you can fit new platforms under the existing Chinatown, you can probably fit station-less tunnels too.)

If you have all of Essex St between Washington St and Kingston St to make that climb it's not a problem. What I'm not sure of is is what the depth of the CAT under Surface Road is, and if the line can fit over it. Even if you need to make the whole climb up from ~40-50ft depth to ~20ft(?) depth between Surface Road and Atlantic Ave, that's still not even as steep as the climb from North Station to Science Park, which, while slow, is within the capabilities of existing and future GL rolling stock.

Maybe, but without seeing the list of station sites considered I don't want to jump to that conclusion immediately. I don't think it really wouldn't have made sense with the loop, so that could have just put it out of the running from the start. Or some big-wig demanding two stops could have also done it, or maybe the shorter platforms for buses just made it the better option.

There is definitely a transfer penalty, but I don't think it's that bad. The block is about 500 ft long, let's target a platform length of 300ft. That leaves roughly a 100ft walkway to Chinatown or Boylston Stations on either side. The complexity of navigating the inbound/outbound platforms at Boylston would probably make that transfer longer, but most GL riders would have other, better options such as just transferring at Arlington or Bay Village. Nubian riders would have a longer transfer but even in the worst case scenario that's only going to be as bad as the longest OL/BL connection at State. Definitely not ideal but also not catastrophically bad.

At surface road, it's apparently just barely below the surface, so you'll have to go under.

(Worth noting this section is from 1993, so while accurate for the segment being discussed, at the time they were actually proposing an Ave Lafayette & Avery routing with Chinatown platforms underneath Hayward.)

That said, one additional consideration to be given to configuring and underpinning these new platforms is that LRT doesn't need as much extra volume as buses do. The existing SL and proposed phase 3 tunnels are 15ft (12ft lane with curbs) wide for a 8.5ft wide vehicle because it's an unguided busway - they needed the extra space to accommodate the vagaries of a human operator driving a bus. It's why they had to eliminate the alt of reusing the Pleasant St incline for the portal fairly early. Conversely, as the green line is railbound, it's standards are written such that it only needs 6in of running clearance from walls / other vehicles - as of GLX specs, it had 12' of horizontal clearance preferred, 11' min. As such, much of GLX's corridor was built spaced to 11'6" on center. That cumulative 6 feet saved might ease fitting both platforms on a single level for Chinatown.

 

[TheRatmeister]

Mid-block would've required a 6.4% grade in violation of the maximum tolerances.

But not between Chinatown/Boylston (I'm going to call it Essex) and South Station, but between Essex and the portal at Tufts Medical Center. A portal which would not reuse the one under Eliot Norton Park, but would involve making a very sharp turn onto Washington St, or some other combination of crazy tight turns and sharp inclines. The Eliot Norton alt. seems to have been essentially discarded because that would require removal of the park and they didn't feel like discussing that further. If we're being less obstinate, we could discuss turning nearby surface parking lots into new park space, or decking over part of the Pike for new park space to mitigate this removal. You know, problem solving. However, it still seems possible that the curve connecting the Tremont St Subway to the Essex St subway is not feasible, in which case the Seaport-Huntington OSR would not be available without using the existing Huntington Ave Subway connection at Copley, which is undesirable for flat junction reasons. I wouldn't consider that to be a strict deal-breaker but it would undeniably be a bummer and negatively weigh on the value proposition.

The egresses would've been more constrained making it less able to cope with the dwells of the increased ridership from pooling the two stations,

In terms of passenger demand, Boylston/Chinatown combined are less busy than Copley, so I find that argument less than convincing, especially with larger vehicles and longer platforms. Boylston probably needs to be completely rebuilt no matter what so a capacity increase is hardly out of the question.

The egresses would've been more constrained making it less able to cope with the dwells

It's less able to cope with dwells because the proposed headway between vehicles as part of SL Phase 3 is 50 seconds, 72 BPH. If we cut that back to more like 3 minute headways, I'm much less concerned.

 

[F-Line to Dudley]

That said, one additional consideration to be given to configuring and underpinning these new platforms is that LRT doesn't need as much extra volume as buses do. The existing SL and proposed phase 3 tunnels are 15ft (12ft lane with curbs) wide for a 8.5ft wide vehicle because it's an unguided busway - they needed the extra space to accommodate the vagaries of a human operator driving a bus. It's why they had to eliminate the alt of reusing the Pleasant St incline for the portal fairly early. Conversely, as the green line is railbound, it's standards are written such that it only needs 6in of running clearance from walls / other vehicles - as of GLX specs, it had 12' of horizontal clearance preferred, 11' min. As such, much of GLX's corridor was built spaced to 11'6" on center. That cumulative 6 feet saved might ease fitting both platforms on a single level for Chinatown.

That's probably not going to be possible. The '05 EIS frustratingly doesn't include a total width measurement for stacked Chinatown, but it does offer one for almost identically stacked Boylston: 45'4" from outermost concrete wall pour to outermost concrete wall pour (meaning, subtract at least a few total feet's worth of concrete to net playable space). At both stations the egresses were contained within the pour without jutting outside because of abutting building foundations, which likely constrains you to a center-island platform because of the need to consolidate the egresses instead of having double on each side. You're allowed a minimum of 6 feet platform width before the nearest obstruction (seat, signpost, trash can, etc.), with the T coalescing on a 9-ft. minimum for new-construction side platforms. Center islands with egresses in the middle typically run 20 ft. (East Somerville) to 25 ft. (Union Square, most Orange Line-north stops) to as large as 35 ft. (Lechmere, Wollaston, most Orange Line-south stops), and of course they'd need to be larger the busier and more transfer-oriented a station is.

Take the walls, the tracks, the egresses, and the minimum-most platform allowances and you're probably scraping the limit. Take a platform allowance worthy and more or less required of a busy transfer station, and you're easily exceeding the limit. I'm reasonably confident that even an LRT do-over would be required to do stacked platforms at Chinatown for the platform accommodations to be in any way worth their salt at moving people.

 

[Riverside]

At surface road, it's apparently just barely below the surface, so you'll have to go under.
View attachment 60850
(Worth noting this section is from 1993, so while accurate for the segment being discussed, at the time they were actually proposing an Ave Lafayette & Avery routing with Chinatown platforms underneath Hayward.)

That said, one additional consideration to be given to configuring and underpinning these new platforms is that LRT doesn't need as much extra volume as buses do. The existing SL and proposed phase 3 tunnels are 15ft (12ft lane with curbs) wide for a 8.5ft wide vehicle because it's an unguided busway - they needed the extra space to accommodate the vagaries of a human operator driving a bus. It's why they had to eliminate the alt of reusing the Pleasant St incline for the portal fairly early. Conversely, as the green line is railbound, it's standards are written such that it only needs 6in of running clearance from walls / other vehicles - as of GLX specs, it had 12' of horizontal clearance preferred, 11' min. As such, much of GLX's corridor was built spaced to 11'6" on center. That cumulative 6 feet saved might ease fitting both platforms on a single level for Chinatown.

@Stlin, thanks for digging this out! If you have the link to the 1993 doc handy, I'd love to look. (I found one document from '93, but I think this may have come from the appendix, which wasn't included.)

@Teban54, I think this might actually answer our questions about Lincoln/Essex/Surface Road: if I'm reading this right, the Transitway tunnel would have just fit under the highway... which means that space under the highway is available for the curve:

My concern had been about fitting the curve between the Lincoln Plaza building and the wall of the tunnel (broadly indicated in dark red above). But if it's feasible to use the space underneath the tunnel, then you have this wide open area (blue highlight) to work in, allowing you to "swing wide" (illustrative range indicated with yellow hashing) for the curve. And, as I've tried to indicate, you don't actually need to swing that wide -- we just need a little bit of wiggle room to ensure a reasonable curve radius.

But, if that curve works, then I think there's a pretty viable plan overall there. Marginal Road -> highway spaghetti -> Lincoln St -> Essex -- that's definitely not bad.

Post Office Square Redux/Essex St Subway etc: I know I said this already but I think it does bear repeating: in order to use all four tracks at Boylston, the point where your "Post Office Square" services branch off also needs to be (broadly) where your Huntington services join the Boylston Street Subway. If your divergence point is somewhere around Charles St (i.e. the extant provision for the Post Office Square extension), then you also need to build a way for Huntington services to get from Prudential to Charles St, which balloons the scope significantly. Criss Cross via Lincoln St limits tunneling to Copley Square proper plus one block south, Lincoln & Essex Streets (above), and alongside the Mass Pike (where impact will be less than almost anywhere else).

For comparison, here's how much of the "neighborhood street grid" would need to be torn up for different segments:

• Lincon + Essex: 2,000 feet
• Copley (plus a block on Dartmouth and Huntington each): 2,000 feet
• Tremont St between Marginal Road and Elliot Norton Park (needed for Tripod): 450 feet
• Charles St + Stuart St + Kneeland St + Lincoln St + Essex St (for Kenmore <> Seaport): 4,300 feet
• Stuart St + Columbus + Park Plaza: 3,500 feet

I'm not counting the segments along the Mass Pike because they are significantly less disruptive (especially since a large fraction of the segment along the Mass Pike actually has no direct abutters to the north).

Criss Cross would be 4,000 feet of tunneling, plus maybe another 450 if we want the extra flexibility of that connection. Using Stuart + Kneeland is almost twice that, at 7,800 feet (and along major roads). For comparison, Herald St to Nubian Square along Washington St is about 8,500 feet.

Insofar as political capital for disruptive tunneling is both finite and fungible, I'd rather spend that capital on something like the Nubian Subway.

 

[Teban54]

I'll get to the (excellent discussions of) SL Phase 3 engineering details at a future date, but just a quick nitpick on a particular point:

For comparison, here's how much of the "neighborhood street grid" would need to be torn up for different segments:

• Lincon + Essex: 2,000 feet
• Copley (plus a block on Dartmouth and Huntington each): 2,000 feet
• Tremont St between Marginal Road and Elliot Norton Park (needed for Tripod): 450 feet
• Charles St + Stuart St + Kneeland St + Lincoln St + Essex St (for Kenmore <> Seaport): 4,300 feet
• Stuart St + Columbus + Park Plaza: 3,500 feet

I'm not counting the segments along the Mass Pike because they are significantly less disruptive (especially since a large fraction of the segment along the Mass Pike actually has no direct abutters to the north).

View attachment 60855

Criss Cross would be 4,000 feet of tunneling, plus maybe another 450 if we want the extra flexibility of that connection. Using Stuart + Kneeland is almost twice that, at 7,800 feet (and along major roads). For comparison, Herald St to Nubian Square along Washington St is about 8,500 feet.

Insofar as political capital for disruptive tunneling is both finite and fungible, I'd rather spend that capital on something like the Nubian Subway.

An implicit assumption of your "4,000 vs. 7,800" is that you get the Marginal Rd subway for free (Back Bay - the Highway Interchange District), or that it's already built.

Technically -- I repeat, technically -- the Marginal Rd subway doesn't need to be built. One can imagine a Criss-Cross or even 2x2 setup that mostly tunnels under Stuart-Kneeland:

• Huntington - Park Inner: Uses Stuart St - Columbus Ave - Boylston St (your teal line)
• Kenmore - Seaport: Uses Boylston St - Charles St S - Stuart-Kneeland Sts (your light green line)
• (Optional) Huntington - Seaport: Straight on Stuart-Kneeland Sts
• (Optional) Kenmore - Park Inner: Straight on Boylston St (Central Subway)

There are multiple issues with this plan (most notably: likely greater tunneling cost, surface disruption, engineering challenges of the interchange, and difficulty of placing transfer stations). But it has some distinct benefits that are sometimes underappreciated:

• Shorter tunnel distances overall, if Marginal Rd subway is not built
• Shorter route for Kenmore-Seaport trains, due to much less "southward deviation"
• The route feels more "central" than following the Pike. Obviously it depends on station locations, but the entire Stuart-Kneeland alignment is much closer to jobs, shops, restaurants and hospitals than Marginal Rd.

With that said, I do think that having Kenmore-Seaport trains run Copley-BBY-Marginal will still turn out better, mostly due to easier and less disruptive tunneling. But involving Stuart-Kneeland in some capacity has its own merits that are worth considering.

 

[Stlin]

@Stlin, thanks for digging this out! If you have the link to the 1993 doc handy, I'd love to look. (I found one document from '93, but I think this may have come from the appendix, which wasn't included.)

Sure!

There's a couple of these; they're identically titled and dated, but are in fact different.

Less detail:

South Boston Piers/Fort Point Channel Transit Project

books.google.com

More detail:

South Boston Piers/Fort Point Channel Transit Project

books.google.com