Green Line Reconfiguration

[F-Line to Dudley]

Java King found quite a detailed document on proposals for a Green Line to Nubian (and Mattapan), in his post here.

It's got the Urban Ring, too.

 

[Riverside]

@Teban54, thank you for your thoughtful comments and kind words! I'm sorry it's taken me so long to respond.

Overall, I should clarify that these cost estimates are extremely provisional. If we're thinking in terms of "significant figures", this estimate probably has only one sig fig, meaning the conclusion here is that the cost is 1 x 10^10 -- something akin to "ten billion, give or take ten billion", or, more practically, "between 5 and 20 billion". There's a lot of variability and a lot of uncertainty, and lots to debate (as we'll see below).

But overall I don't think that the variability and debate fundamentally changes the "between 5 and 20 billion" number, and I think the same is true for the NSRL -- which is why I think these projects are comparable and in fact should be compared.

Riverside said:
For example, the Commonwealth Subway would be highly valuable for increasing speed and reliability, but isn’t critical to any particular project.

I actually think that in a full-build GLR, the Commonwealth Subway will be more important than shown here. Depending on build configuration, up to 4 branches may traverse the stretch at what's now BU Central: Boston College, Oak Square, Harvard, and Grand Junction. That's the same number of branches as present-day Copley-GC (which is already a bottleneck on today's Green Line), but even worse, running at street level for 0.7 miles with no grade separation and possibly no signal priority.

To make things worse, each of the 4 branches will likely involve some street-running, with the possible exception of Harvard. Especially BC/Oak Sq, and especially if Grand Junction service originates all the way from Chelsea or Airport. That's a huge reliability concern that may be even worse than today's Green Line (since the D has full grade separation).

You raise compelling points about how "beefy" of a Comm Ave subway is needed to make this whole thing work. I would argue we can simplify things a bit, though:

Up to 4 branches may traverse the stretch [between Kenmore and BU Bridge]: Boston College, Oak Square, Harvard, and Grand Junction.

I think all four of these branches being routed over this stretch is both unlikely and undesirable.

Sending a Grand Junction branch to Kenmore solves no problems and creates new ones: Kenmore station cannot terminate trains from Comm Ave without a significant/total rebuild, and is not itself an employment or residential destination -- both BU and Longwood are much larger hubs. Sending a Grand Junction branch around beyond Kenmore consumes capacity in the Boylston St Subway, and provides poorer Cambridge <> Back Bay and Cambridge <> Downtown service compared to a Grand Junction/Lechmere wraparound, the Red Line, or something via Mass Ave. (Mass Ave @ Grand Junction <> Copley is 1.5 miles via Mass Ave + Green, and is nearly twice that via BU Bridge.) If we can find a way to send a Grand Junction line into the Beacon St subway to access the Kenmore Loop that way, then I think it's fine. But I don't think it's worth spending significant money to enable a Grand Junction/Comm Ave wraparound.

A Harvard branch has a few different potential purposes:

• Urban Ring service between Harvard and something in the southwest quadrant (ideally Longwood)
• Urban Ring service between Harvard and West Station (for Porter-style transfers)
• Radial service from West Station to Back Bay/Downtown
• Radial service from Allston to Downtown

The Urban Ring services can't terminate at Kenmore, so a Comm Ave subway is less useful to them. Radial service from West Station is valuable, but could be achieved with strong Regional Rail frequencies (and in the long term I'd argue would be a strong western anchor for Blue Line service). That leaves a OSR between Lower Allston and Downtown as the unique benefit of running a Harvard branch via a Comm Ave subway. That's a worthwhile benefit, but I think it probably is comparable to the benefits of a restored Oak Sq branch, so I think we'd see either an Oak Square branch via a Comm Ave subway, or a Harvard branch, but not both.

Which is why I think it's more likely that we will see a maximum of two, not four, services between Kenmore and BU Bridge. If one of those branches is a greenfield LRT line to Harvard (especially a grade-separated one), then that might tip the scales in favor of a subway, but I don’t think it’s a sure thing.

Overall, the potential Grand Junction and Harvard branches benefit much more from a connection toward Longwood rather than Kenmore. If Fenway station becomes the “landing target”, then we’re also talking about a slightly shorter corridor as well — about 2600 feet vs about 3600 feet to Kenmore. So I think we get more “bang for buck” by focusing on building that connection, rather than spending the money on a Comm Ave subway whose services still will be impacted by street-running reliability. The Kenmore Division is always going to mix poorly with anything new we build, and I think that holds true here.

I agree that a 4 track subway can mitigate that, but where will those extra two tracks go after Comm Ave? I suppose you can build a new Kenmore Under terminal, but again, that's a lot of money to spend for a service that isn't even where people need to go. If we want to build a Comm Ave subway, I'd rather build it as an extension of the Blue Line to a short-term terminal at BU Bridge, staged to eventually extend to West Station. That would bring both the Urban Ring transfer and the Blue Line terminal to where the employment center actually is: Boston University.

Another consideration here is the BU Bridge junction. You need the junction to enable services to Packards Corner, West Station and Grand Junction. A surface intersection at the current BU Bridge junction doesn't make the cut to me given the existing traffic patterns, and a viaduct may have difficulty descending onto the BU Bridge ROW. So you need a tunnel for the junction even in a smaller build - and if that's the case, it may be an additional argument for just building the full Commonwealth Subway once, instead of in two stages (and disrupting traffic and B branch operations twice).

As I've alluded to above, I think the Aldgate junction at BU Bridge only needs to point in three directions: Cambridgeport, West Station, and one direction south (e.g. Park Dr). Packards Corner doesn't need service from anywhere other than Kenmore, and as I've argued above, I think Grand Junction <> Kenmore is a poor investment. (A non-revenue connection could be operationally valuable, although I don't think it would be make-or-break.) In some ways, I'd rather settle on a design for the Urban Ring stuff first, and then add a Comm Ave subway as a secondary priority.

But yeah, either way, the Comm Ave subway is "only" $1.4B, which is practically a rounding error in this exercise. I definitely think it's an important question to resolve, but there are many important questions to resolve for the Green Line Reconfiguration overall, and I'm not trying to suggest I have all the answers so far!

I'm assuming you're using a street-running route via Mountfort St and Park Dr. It can work in a minimum build, but I'm concerned about reliability, given traffic, several signalized intersections, and the low likelihood of getting any dedicated ROW. This is in addition to adding even more complexity to the BU Bridge junction - now you have two possible directions on the southeast side.

I was trying to leave this corridor open-ended on exact route/alignment, because I agree there are problems with using Park Dr, though probably I should've erred on the side of more expensive estimates, and treated it like a "tunnel" segment. At "tunnel" costs (which probably are also a vaguely decent estimate for elevated costs), it's a little less than $1B. But, as mentioned above, to me it'd be a no-brainer to reallocate those Comm Ave subway costs to build a connector like this.

(And, assuming we can leverage Mass Pike air rights to avoid actually running at-grade along Mountfort, I don't actually think we need to be quite so pessimistic about Park Dr. Close the northernmost block [between Mountfort and Buswell] to public traffic, reclaim the western half of the street between Buswell and Fenway station [easier said than done, true], and traverse one traffic light. That's not ideal, but that doesn't seem impossible to me.)

 

[Riverside]

We'll see how ridership evolves after SL3 extension and SL6 are implemented, but if they're huge successes, it may be an argument in favor of LRT conversion. LRT does miss out on Glendale and Upper Broadway in Everett (and has less convenient pedestrian access from Sweetser Circle), but the Silver Line alternatives analysis mentioned that the Glendale-Kendall alternative and Chelsea-Kendall alternative have similar projected ridership, so there's good potential for LRT and BRT to complement each other on this corridor.

Yeah I wasn't super clear here, I didn't really mean to suggest that SL3x/SL6 would make LRT appear less necessary, but more that we will have a more precise understanding of the needs along this corridor. (And I continue to maintain that the SL3x/SL6 numbers need to be understood in the context of the current utterly awful connectivity between Sullivan and Kendall where literally your only option is the CT2, which has garbage headways of 20min peak/60 min off-peak. Chelsea LRT may still be required even with Sullivan <> Kendall LRT, but the justification will very likely look different.)

I know you probably didn't include any southside Urban Ring infrastructure towards the GLR cost estimations (e.g. Nubian to Huntington), but I think it might be a good time to revisit The EGE's idea of a D-E connector via Longwood instead of Brookline Village.

One thing you illustrated well here is how "expensive" a subway D-E connector is - about 0.9 miles from LMA station to Brookline Village. Tunneling via Longwood Ave (or Francis St but I'm doubtful on that) is even shorter on paper at 0.7 miles. Of course, a Longwood subway will be much more expensive due to building mitigations, but I think this is where a full cost-benefit analysis is needed.

I agree that it is good to periodically revisit the question of a Longwood subway. And to your point, using the very coarse measurements I'm using here, the tunnel mileage is likely shorter via Longwood proper anyway. One thing to note, however, is that (as @The EGE noted on his map) you need to find someplace to terminate your Urban Ring services; I'm using the Fenway Branch to Brookline Village to do that, but a Longwood subway would probably require provisions for terminating operations at the "LMA Central" station (unless you want to introduce a capacity pinch by interlining circumferential Urban Ring services with radial Huntington services).

That being said, I think it's highly likely that my estimation method vastly overestimates the cost of a Huntington subway. My numbers essentially assume deep bore tunneling for everything; of the 1.7 miles between Northeastern and Brookline Village, over half of it (0.9 miles) is under an 80-year-old reservation, on a street that is extremely wide (i.e. could see partially closure for C&C with less acute impact to surface traffic). The next 0.4 miles (to South Huntington) are under a slightly narrower street that is still extremely wide. So it's easily possible that 1.3 of the 1.7 miles (76%) could be handled through cut-and-cover. That last 24% is definitely a wildcard, but that's still a much smaller fraction.

That all being said: a Huntington Subway with a D-E Connector and a connection to the Pleasant St Portal is the heart of this whole shebang. To me, easily the top priority after the Blue-Red Connector. It makes everything else possible, and would be transformative in its own right. (And, again, to me seems like a much better use of $5B than building half of the NSRL.)

 

[Teban54]

Thanks for the detailed reply, @Riverside !

Re: BU and Commonwealth Ave Subway

I think there's a bit of confusion arising from my somewhat subjective initial comment, so let me clarify things:

I do agree there are mainly two trunks in this corner of the system:

1. Oak Sq and Boston College - Comm Ave and BU - Kenmore (a more traditional streetcar route)
2. Grand Junction and West Station - Somewhere around BU - Highland Branch or LMA (or, in the worst case, Kenmore) (a more Urban Ring-oriented route)

My comment about possibly having 4 branches through Comm Ave was envisioning the worst case, where Trunk 2 also uses the same ROW as Track 1 along Comm Ave (surface or subway) through part of the route, regardless of whether it goes to Kenmore. I've been brainstorming my own proposal for Trunk 2, which I'll hopefully post at some point, but my current thought involves also going under Comm Ave and turning south just before Kenmore. I think there are a few reasons why having Trunk 2 use Comm Ave may be desirable, the biggest of which being that it serves BU directly - I'd argue that Urban Ring users need access to BU just as much as Blue Line riders.

Of course, this is not the only way for Trunk 2 to play out, and with any other alignment that doesn't involve using the same ROW along Comm Ave as Trunk 1, there wouldn't be a huge need for a Comm Ave subway at all (in which case it would be only for Trunk 1). Thus, I agree that the biggest priority here is to build whatever infrastructure that allows Trunk 2, and in some but not all cases, it may end up being a Comm Ave subway that also allows Trunk 1 to benefit.

Minor notes:

Kenmore station cannot terminate trains from Comm Ave without a significant/total rebuild, and is not itself an employment or residential destination -- both BU and Longwood are much larger hubs.

As I've recently learned myself, there are indeed a few academic buildings right at Kenmore. Some buildings may also be closer to Kenmore than a combined BU station with a Comm Ave subway or BLX. And of course, an option to transfer to the B or the 57 is always there. BU East and BU Central stations are still the biggest destinations, and should be prioritized as much as possible; but if Kenmore is the best we can do (for both Urban Ring and BLX), it may be workable.

Also, due to how stretched out BU is, very few alignments (especially for Urban Ring) can offer perfect coverage. Employees working at the gyms and other student life facilities will want to go to Amory St and Babcock St stations, for example.

(And, assuming we can leverage Mass Pike air rights to avoid actually running at-grade along Mountfort, I don't actually think we need to be quite so pessimistic about Park Dr. Close the northernmost block [between Mountfort and Buswell] to public traffic, reclaim the western half of the street between Buswell and Fenway station [easier said than done, true], and traverse one traffic light. That's not ideal, but that doesn't seem impossible to me.)

From the BU Master Plan (which someone posted here a while ago), BU does have air rights over Mass Pike, and plans to put a few buildings over it. Realistically, a street-running approach may end up having to run on Mountfort.

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

Re: D-E connector and Huntington Subway

I agree that Huntington Subway up to Brigham Circle is an absolute must, and is likely the cheapest among all tunneling that needs to be done for GLR. My intent was more about comparing the length of a Longwood Ave subway vs Brigham Circle-Brookline Village, or LMA Station-Brookline Village, making the options a lot more comparable in length alone. Although Longwood Ave subway will likely be much tricker and costlier to build than Brigham-BV, even if the latter requires some amount of TBM to cross the river.

My idealized crayon map would feature an LMA subway, a Huntington subway that opens up to street level at Brigham Circle for the E, and a Ruggles St subway (or portal) that connects to the southside Urban Ring. In this way, Ruggles becomes an ideal terminal for northside Urban Ring services, allowing full frequencies on the important Ruggles-LMA connection.

A build sequence that works best with this proposal would be:

• Phase 1: Huntington subway up to Brigham Circle, and a D-E connector fully on the surface.
  • D trains (or Needham trains only) use D-E surface connector with street running between Brookline Village and Brigham Circle - may have political issues
  • North UR trains terminate at Brookline Village
• Phase 2: Longwood Ave subway, and a short Ruggles St subway or portal that connects to OL
  • D trains run via Longwood and Huntington, with full grade separation again
  • North UR trains terminate at Ruggles via Longwood
• Phase 3: Southside UR
  • North UR trains likely still need to terminate at Ruggles
  • South UR trains terminate at BU, Kendall/West Station, or even Sullivan/Harvard

However... This is very idealized. In reality, I'm expecting a subway D-E connector via BV, with a Longwood subway becoming its own project if we're lucky (since it will essentially end up as just the UR, with a lot less benefits).

Minor notes:

a Longwood subway would probably require provisions for terminating operations at the "LMA Central" station (unless you want to introduce a capacity pinch by interlining circumferential Urban Ring services with radial Huntington services).

Given how narrow Longwood Ave is, I doubt there will be space for terminating operations. And a Longwood subway will likely interline with Huntington services anyway, since in a build similar to The EGE's, it will serve both Riverside-LMA-Symphony trains and GJ-LMA-Ruggles trains. There will also likely be no space at the Huntington/Longwood intersection, so it's either a flat junction (welp), or stacked tracks for both Huntington subway and Longwood subway here. (You may end up needing to stack the tracks for Longwood subway anyway given its width.)

So a Brookline Village alignment for UR is significantly less complicated operations-wise - but only if you completely disconnect the two halves of UR and isolate them from Huntington subway. This means terminating northside UR services at Brookline Village, and making southside UR either BRT or LRT that terminates without crossing Huntington. The moment you want to run some zigzag patterns via Brookline Village to connect the two halves, you're interlining with Huntington services again.

I can definitely see reasons to do that, since few passengers will want to go through the BV zigzag... But let's just say it's a lot less elegant.

(Honestly, it's just hard to come up with proposals for the southern half of Urban Ring from LMA to Seaport and even to Airport, especially compared to the northern half. Perhaps BRT is what we'll eventually have to settle on, but I worry about ridership and reliability, especially given the ongoing discussion about LRT vs BRT ridership for the Silver Line.)

 

[Riverside]

Re: BU and Commonwealth Ave Subway

I think there's a bit of confusion arising from my somewhat subjective initial comment, so let me clarify things:

I do agree there are mainly two trunks in this corner of the system:

1. Oak Sq and Boston College - Comm Ave and BU - Kenmore (a more traditional streetcar route)
2. Grand Junction and West Station - Somewhere around BU - Highland Branch or LMA (or, in the worst case, Kenmore) (a more Urban Ring-oriented route)

My comment about possibly having 4 branches through Comm Ave was envisioning the worst case, where Trunk 2 also uses the same ROW as Track 1 along Comm Ave (surface or subway) through part of the route, regardless of whether it goes to Kenmore. I've been brainstorming my own proposal for Trunk 2, which I'll hopefully post at some point, but my current thought involves also going under Comm Ave and turning south just before Kenmore. I think there are a few reasons why having Trunk 2 use Comm Ave may be desirable, the biggest of which being that it serves BU directly - I'd argue that Urban Ring users need access to BU just as much as Blue Line riders.

Of course, this is not the only way for Trunk 2 to play out, and with any other alignment that doesn't involve using the same ROW along Comm Ave as Trunk 1, there wouldn't be a huge need for a Comm Ave subway at all (in which case it would be only for Trunk 1). Thus, I agree that the biggest priority here is to build whatever infrastructure that allows Trunk 2, and in some but not all cases, it may end up being a Comm Ave subway that also allows Trunk 1 to benefit.

Gotcha, yeah this makes a lot of sense and I think I agree with all of it. That's an interesting idea to have a "zig-zag" that runs east on Comm Ave until (e.g.) Blandford, cuts under the Pike via TBM (including a Lansdowne subway station?), and hooks into the Fenway Branch clear of Beacon Junction. (Probably has been thrown around up thread, though I can't call it to mind right now.)

I need to sit down and diagram it out to be sure, but I suspect that, due to one elevation change or another, the Trunk 1 Radial tunnel would need to be parallel to but separate from the Trunk 2 Circumferential tunnel. The Radial tunnel really just needs to be 1 level below surface, but the Circumferential tunnel will need to be low enough to make it underneath both the Mass Pike and the C Line -- the distance from Comm Ave to the MassPike is 400ish feet, which might be enough to drop one level, but if you're near Blandford, then that's where the C Line passes under the MassPike, meaning the Circumferential tunnel needs to be at Level -3. (And a parallel but separate tunnel would also allow you to avoid needing a flying junction.) And at the other end (BU Bridge) you've got more spaghetti.

For funsies, here's the real crazy transit pitch: 6 tracks under Comm Ave:

Not labeled in the above, but at this eastern end of the Comm Ave Supersubway, Green and Blue are at about -1, basically as deep as Kenmore. At this point, Gold is probably at -2, so that it can drop to -3 by the time it hits the C Line under the Pike.

While at this end, it's the Green (and maybe the Blue, depends on how we actually want to get to West Station) that needs to drop down under the Pike, while the Gold needs to stay shallow in order to reach a portal and interface with the BU Bridge junction.

But anyway, this is planning contingencies on contingencies. I agree with how you put it: the biggest priority here is to build whatever infrastructure that allows Trunk 2, and in some but not all cases, it may end up being a Comm Ave subway that also allows Trunk 1 to benefit.

From the BU Master Plan (which someone posted here a while ago), BU does have air rights over Mass Pike, and plans to put a few buildings over it. Realistically, a street-running approach may end up having to run on Mountfort.

Yeah, if the Mass Pike isn't available then the Park Dr surface option likely isn't viable -- I think that short surface segment could be swung, but unless Mountfort is closed to traffic it really isn't viable. I increasingly like your idea of tunneling across near Blandford -- make the bored segment as short as possible, leverage the wide ROW of Comm Ave, and take advantage of the Fenway Branch as a landing point.

Re: D-E connector and Huntington Subway

Good thoughts with all of these including the build sequences.

So a Brookline Village alignment for UR is significantly less complicated operations-wise - but only if you completely disconnect the two halves of UR and isolate them from Huntington subway. This means terminating northside UR services at Brookline Village, and making southside UR either BRT or LRT that terminates without crossing Huntington. The moment you want to run some zigzag patterns via Brookline Village to connect the two halves, you're interlining with Huntington services again.

I can definitely see reasons to do that, since few passengers will want to go through the BV zigzag... But let's just say it's a lot less elegant

Yes, this is what I've landed on: terminating (northside) UR services at Brookline Village simplifies ops and maximizes capacity. And yes -- any zigzag via Brookline Village (or any stretch along Huntington -- e.g. between Longwood Medical Area station and MFA station before turning to Ruggles) interlines with Huntington services, and I think that would have an outsized impact on capacity and reliability for a Huntington Subway (which otherwise could be completely isolated from surface services). Given that it would be an inelegant indirect zig-zag service anyway, the interlining doesn't seem worth that cost.

I think it's inevitable that any solution here will be inelegant. I went through some ideas in the Deity Mode thread back in January and one thing I think it highlights is that the physical locations of the key nodes (Nubian, Ruggles, LMA, Brigham Circle, the Fenway Branch) really just make for an impossible circle to square unless we go full TBM.

I still think I land on something like this:

Let a Radial BRT network handle the zig-zag via Ruggles (and leverage Brookline Ave bus lanes to boot), and let a straight shot Circumferential LRT route (ideally cut and cover subway) handle the transfers between Green, Orange, and Nubian. Yes, you are interlining with Huntington services still, but it would likely be a "closed network" of Huntington + Washington + UR services, which is more self-contained.

Ironically, I think the lynchpin to all this is easily overlooked: a infill station here on the Fenway Branch would provide better access to Brigham and the other hospitals at the southern end of the LMA than the current Longwood station does. Obviously it's not as ideal as a subway station in the heart of the LMA, but it would stick to the existing ROW, and for commuters from Harvard, Kendall, and Sullivan, it would get you quite close to the job-heaviest section of Longwood.

But yeah, this...

(Honestly, it's just hard to come up with proposals for the southern half of Urban Ring from LMA to Seaport and even to Airport, especially compared to the northern half. Perhaps BRT is what we'll eventually have to settle on, but I worry about ridership and reliability, especially given the ongoing discussion about LRT vs BRT ridership for the Silver Line.)

...is just very true. It's either expensive or inelegant -- pick your poison.

 

[Riverside]

Gotcha, yeah this makes a lot of sense and I think I agree with all of it. That's an interesting idea to have a "zig-zag" that runs east on Comm Ave until (e.g.) Blandford, cuts under the Pike via TBM (including a Lansdowne subway station?), and hooks into the Fenway Branch clear of Beacon Junction. (Probably has been thrown around up thread, though I can't call it to mind right now.)

Found it. It was an idea from @vanshnookenraggen that is pretty similar to what we're discussing now. He argued in favor of a D-E connector under the Brookline Avenue Playground as a way to make a "wraparound" from the current Longwood station to a subway station at Brigham Circle more palatable:

While I think you would need an additional station around Parker Hill Ave, this looks somewhat feasible and potentially is an equal amount of bored tunneling as would be via a more southerly alignment along Route 9.

If entirely in a dedicated ROW with decent speeds, mayyyybe this could make a zig zag work.

 

[Teban54]

A question we should consider is: Is there a need for a zigzag service, or more accurately, a complete SW Urban Ring that doesn't split northside and southside services in segments?

Thinking about it again, I can definitely see some demand for it. I think the main service pattern here is: Ruggles and Nubian - BU and Kendall.

Both Ruggles and Nubian are major bus hubs, and the former is also a major Commuter Rail transfer station. Both BU and Kendall (plus MIT) are major employment centers. Yet, even in a post-BNRD world, the only reliable bus routes that's remotely close to an OSR between any of the origin-destination pairs are: T1 from Nubian to MIT, T28 from Nubian to Kenmore, and T47 from Ruggles to bus stops in the BU Bridge vicinity. All of them are just a little bit away from where most jobs actually are. The 85 gives a direct Ruggles-Kendall OSR, but it's an infrequent route with 30-min peak frequency.

Commuters that live along the T28 do get two-seat rides to BU by transferring at Kenmore. Commuter Rail riders can transfer at South Station for a two-seat ride to Kendall, but that takes up capacity downtown - and one of Urban Ring's main objectives is to prevent or at least mitigate that. Aside from those, for anyone starting their journey from a Commuter Rail stop or a bus route south of Nubian, they need 3-seat rides (some of which still involve two surface routes or a downtown transfer), or in some cases, 2-seat rides plus walking.

A successful zigzag Urban Ring route turns all these into 2-seat rides from each commuter's origins, or 1-seat rides from Ruggles and Nubian.

I suppose the question becomes:

• Is the demand enough to justify the cost (particularly if Ruggles and Nubian can't both be served with a single UR service pattern, or if the infrastructure can't be combined with a D-E connector)?
• Is the OSR's travel time to BU and Kendall competitive enough? (Especially compared to CR-South Station-RL-Kendall)
  • My opinion is that, even if the travel time is longer but not too bad, it will still draw enough non-CR commuters to switch to the OSR due to convenience

Note that, as I've shown before, a UR station at Washington/Melnea Cass (plus Ruggles) does almost as well as Nubian proper for capturing the Nubian bus routes.

I also intentionally omitted Harvard from the discussion. Even though a Harvard-Nubian Urban Ring OSR is quite doable, a downtown transfer to the Red Line becomes more attractive here, due to the longer distance and the UR Harvard branch possibly involving street running.

 

[Riverside]

A question we should consider is: Is there a need for a zigzag service, or more accurately, a complete SW Urban Ring that doesn't split northside and southside services in segments?

Thinking about it again, I can definitely see some demand for it. I think the main service pattern here is: Ruggles and Nubian - BU and Kendall.

Both Ruggles and Nubian are major bus hubs, and the former is also a major Commuter Rail transfer station. Both BU and Kendall (plus MIT) are major employment centers. Yet, even in a post-BNRD world, the only reliable bus routes that's remotely close to an OSR between any of the origin-destination pairs are: T1 from Nubian to MIT, T28 from Nubian to Kenmore, and T47 from Ruggles to bus stops in the BU Bridge vicinity. All of them are just a little bit away from where most jobs actually are. The 85 gives a direct Ruggles-Kendall OSR, but it's an infrequent route with 30-min peak frequency.

Commuters that live along the T28 do get two-seat rides to BU by transferring at Kenmore. Commuter Rail riders can transfer at South Station for a two-seat ride to Kendall, but that takes up capacity downtown - and one of Urban Ring's main objectives is to prevent or at least mitigate that. Aside from those, for anyone starting their journey from a Commuter Rail stop or a bus route south of Nubian, they need 3-seat rides (some of which still involve two surface routes or a downtown transfer), or in some cases, 2-seat rides plus walking.

A successful zigzag Urban Ring route turns all these into 2-seat rides from each commuter's origins, or 1-seat rides from Ruggles and Nubian.

I suppose the question becomes:

• Is the demand enough to justify the cost (particularly if Ruggles and Nubian can't both be served with a single UR service pattern, or if the infrastructure can't be combined with a D-E connector)?
• Is the OSR's travel time to BU and Kendall competitive enough? (Especially compared to CR-South Station-RL-Kendall)
  • My opinion is that, even if the travel time is longer but not too bad, it will still draw enough non-CR commuters to switch to the OSR due to convenience

Note that, as I've shown before, a UR station at Washington/Melnea Cass (plus Ruggles) does almost as well as Nubian proper for capturing the Nubian bus routes.

I also intentionally omitted Harvard from the discussion. Even though a Harvard-Nubian Urban Ring OSR is quite doable, a downtown transfer to the Red Line becomes more attractive here, due to the longer distance and the UR Harvard branch possibly involving street running.

Good analysis. Assorted thoughts/braindump incoming:

Really this is two corridors/destination pairs: RugglesNubian <> BU and RugglesNubian <> Kendall.

For BU, I do think the combination of the T28 and T47 will actually be pretty strong; to the point you made earlier, commutes to BU will likely always have a skew toward needing to slide along Comm Ave one way or another from either Kenmore or BU Central -- the elongated shape of the campus ensures that. That being said, rerouting the T47 to reach Comm Ave via St Mary's St (tight turns) would allow for more centrally located stops. But in general, assuming the T can get travel times down for those routes (which I am cautiously optimistic about), I'm pretty sure those routes will be competitive with an LRT zig-zag (if not better), and would offer front-door service in Longwood to boot.

You raise an interesting point about Kendall. Still, I think if the need is to fill a RugglesNubian <> Kendall gap, then there's a straightforward solution: leverage the BRT infrastructure for the T1 on Mass Ave, and run a route from Kendall down Vassar to Mass Ave and then over to Ruggles (and probably Nubian). Ruggles to Kendall via Mass Ave is about 2.5 miles, compared to going via Longwood/BU Bridge, which is at least 3 miles (via a LMA surface route -- a zig-zag like we've been describing above would be comfortably over 4 miles). If we assume the Mass Ave BRT option runs at 8 mph and the LRT zig-zag runs at 12 mph, then the Mass Ave option is indeed still faster (18 min vs 20 min).

And that comparison is broadly similar for surface BRT vs zig-zag LRT to BU as well, if I recall correctly: they're potentially pretty close in terms of travel time.

Coming south from Kendall to Longwood on the northside, I feel that there's a little more "padding" for a slightly more roundabout route, because you'll have riders coming from farther away -- e.g. Sullivan or Chelsea-- so even if you lose some time through a roundabout, you've saved enough time via the direct OSR from Sullivan to still make it competitive. On the southside, most riders will be boarding at Ruggles or Nubian (i.e. no equivalent to Chelsea) and it's a shorter distance from RugglesNubian to LMA, BU, or Kendall than it is from Sullivan to LMA, so you're more vulnerable to the time costs of a roundabout.

I think there's a valid need you've identified -- particularly RugglesNubian <> Kendall -- but I'm still skeptical that a zig-zag OSR is an effective or efficient solution.

That all being said...

One of the advantages of using the Fenway Branch as the anchor of northside LRT is that you can build an imperfect version that terminates at Brookline Village first and if possible later can add a cross-Longwood tunnel and divert. The infrastructure north of Fenway station doesn't really change either way, so you aren't painting yourself into a corner by building that. And since the Fenway Branch ROW can be used as is, there would be minimal modifications that would "go to waste" if you built a Phase 1 to BV and a Phase 2 subway -- at most, your changes to Brookline Village to support terminating trains would become obsolete.

 

[Riverside]

This is neither here nor there, but I was looking again at the Watertown Branch earlier today. Maybe about a year ago, I became skeptical of the usefulness of a Watertown-Porter-Downtown service -- it looked too roundabout to me, and seemed like it would compete poorly against the combination of the T70 + Red Line (particularly if the T70 got bus lanes).

But I looked again today, and actually I think it would be fine? Assuming 16 mph average (which I think is what the Medford Branch does right now), it would take 30 minutes to go from Watertown Square to Government Center. At 8 mph (comparable to today's C, E, or SL4/5), the T70 would take 30 minutes just to get to Central, and then it's another 9 minutes (plus transfer time) to Park Street.

So... yeah, I think it should be fine?

 

[Teban54]

Assuming 16 mph average (which I think is what the Medford Branch does right now), it would take 30 minutes to go from Watertown Square to Government Center.

This might be a questionable assumption if the western half of the Watertown Branch involves street running.

At 8 mph (comparable to today's C, E, or SL4/5), the T70 would take 30 minutes just to get to Central, and then it's another 9 minutes (plus transfer time) to Park Street.

Two other factors that go against the T70:

• Western Ave and River St near Charles River can often have a lot of traffic. I'm not sure how easy it is to implement dedicated bus lanes in the area.
• Taking the T70 at Central is not very intuitive. Inbound is fine, but outbound T70 boards on Green St, which requires a short walk and crossing Mass Ave from the outbound Red Line platform. I've often seen people getting confused here. It's also a somewhat sketchy area.

 

[ritchiew]

Western Ave and River St near Charles River can often have a lot of traffic. I'm not sure how easy it is to implement dedicated bus lanes in the area.

There are some plans for bus lanes on Western, or even turning it into a dedicated transitway. But you're right, there are some higher traffic areas along Western the plans don't cover (which seems backwards)

https://www.bostonplans.org/getattachment/8ae157a4-3b7c-4b16-9aae-8eab6357a536

 

[Aprehensive_Words]

Another argument in favor of the Watertown branch -- it connects three life science hubs (East Watertown, Alewife and East Cambridge, near enough to Kendall).

 

[Riverside]

Riverside said:
Assuming 16 mph average (which I think is what the Medford Branch does right now), it would take 30 minutes to go from Watertown Square to Government Center.

This might be a questionable assumption if the western half of the Watertown Branch involves street running.

Hmm, yeah that's a good point. Let's do some math:

It'll be about 1 mile (or less) of street-running, between School St and Watertown Square. So that means you have 7 miles of 16 mph, and 1 mile of 8 mph, which yields 26.25 min + 7.5 min = 33.75 minutes.

Which does still put us better than the T70 + Red Line combo, though it is true the margin is narrower.

On the other hand, F-Line's sketch of this suggested stop spacing of about ~2200 feet (and honestly I think you could go either way for/against the Beechwood Ave stop), compared to typically less than half of that on the BCE. And depending how many unsignaled (unlit? un-traffic-lit?) cross-streets you have, you could potentially have relatively long stretches of uninterrupted sealed ROW in the middle of the street. I'm not saying it'll make a revolutionary difference, but we might be able to hope for a slightly higher average speed.

Two other factors that go against the T70:

• Western Ave and River St near Charles River can often have a lot of traffic. I'm not sure how easy it is to implement dedicated bus lanes in the area.
• Taking the T70 at Central is not very intuitive. Inbound is fine, but outbound T70 boards on Green St, which requires a short walk and crossing Mass Ave from the outbound Red Line platform. I've often seen people getting confused here. It's also a somewhat sketchy area.

In theory, these problems could both be solved by judiciously applied paint (whether on roads, signs, etc). The question in both cases is how easily that could be done.

(Also, one interesting fact to ponder: the T70 is going to run to Kendall, not Central, meaning there will be an OSR. The Watertown Branch of course wouldn't be siphoning away Watertown <> Kendall commuters, so, assuming travel times remain speedy, probably this extension would relieve pressure on the Red Line for those going the single stop after transferring at Central. I wonder what kind of shift this will have the dynamics in this corner of the network.)

Another argument in favor of the Watertown branch -- it connects three life science hubs (East Watertown, Alewife and East Cambridge, near enough to Kendall).

Yeah -- I think it makes a lot of sense to run trains to Watertown from Porter. The part I'm fuzzier on is where should the trains go after Porter. Is the route too roundabout to take up slots in the Central Subway? (Until yesterday, I had thought "yes", and now I am pondering.) That's one reason I've considered sending these trains to Sullivan instead, so they can pull double-duty by linking Porter + Union with Sullivan as a circumferential service. (Of course, doing so loses you the OSR to Lechmere. Easy single platform transfer at Union, but still.)

On the other hand, we know that Watertown <> Downtown is a veritable market, while Watertown <> Sullivan is less obvious. And capacity concerns in the Central Subway can be eased in part by operating some/all Watertown trains through the Brattle Loop; the loss of Red transfer at Park St is offset by the availability of a transfer at Porter. Running into Downtown may just be the simplest solution.

So, I dunno. In the medium-term, I think it may be the strongest option for extension beyond Porter (and would offer a stronger rallying cry: "GLX to Watertown" sounds better than "GLX to Porter"). But I wish it were more of an unambiguous slamdunk.

 

[Aprehensive_Words]

The part I'm fuzzier on is where should the trains go after Porter.

This brings up a concern I've started to have about schemes to run the Green Line through Porter: How on earth are you going to connect it to the existing Red Line station, and offer a behind-the-fare-gate transfer?

On the latter, I guess I can accept waving away the question of rotating the line of fare gates 90 degrees or so (assuming the smaller number of gates still has the necessary throughput) courtesy of some yet-to-be-developed universal fare system. Doesn't seem that hard, AFC 2.0's travails aside.

But the former seems a lot stickier. Consider this highly sketchy sketch of the arrangement of the Porter ticket hall and commuter rail tracks:

The CR ROW is too narrow to quad-track things, so you'd need to run the GL trains underneath. But the CR platform is only about 1 story above the ticket hall floor level. It seems like you're left with one of two fairly expensive propositions: A) digging the GL tunnel even deeper, below the floor level of the ticket hall, or b) digging the ticket hall down so its floor level is below the GL tracks. It looks like far more than a simple cut-and-cover job, right? Or does the fact that the GL tracks would effectively be at only a "sub-basement 2" level thanks to the CR tracks already existing in a trench make it less daunting?

 

[Highwayguy]

RE: Porter, instead of going under the CR for the segment where the ROW necks down, what about going elevated above? Seems much less impactful to CR ops during construction in addition to likely being much cheaper. If the GL could cross Beacon St at grade on the road bridge (I’m assuming the existing bridge would have to be demolished and reconstructed to accommodate this), an El wouldn’t even be visually intrusive since it would be at the same elevation of Somerville Ave. Could even cheap out and slap some north station style fare gates to bring the CR platforms into fare control so only vertical circulation between Green/ CR platforms would be needed to access the RL.

 

[bigeman312]

RE: Porter, instead of going under the CR for the segment where the ROW necks down, what about going elevated above? Seems much less impactful to CR ops during construction in addition to likely being much cheaper. If the GL could cross Beacon St at grade on the road bridge (I’m assuming the existing bridge would have to be demolished and reconstructed to accommodate this), an El wouldn’t even be visually intrusive since it would be at the same elevation of Somerville Ave. Could even cheap out and slap some north station style fare gates to bring the CR platforms into fare control so only vertical circulation between Green/ CR platforms would be needed to access the RL.

I love this idea in all ways except one: it makes a future extension beyond Porter much less feasible.

With an indefinite terminus of Porter, though, it works great, in my estimation. No need to fare-gate the Green Line platform, if the other GLX stations are already not fare-gated. You could even build a new headhouse and vertical transportation at Somerville Ave opposite Mossland for better GL and CR access for Spring Hill.

 

[Riverside]

This brings up a concern I've started to have about schemes to run the Green Line through Porter: How on earth are you going to connect it to the existing Red Line station, and offer a behind-the-fare-gate transfer?

On the latter, I guess I can accept waving away the question of rotating the line of fare gates 90 degrees or so (assuming the smaller number of gates still has the necessary throughput) courtesy of some yet-to-be-developed universal fare system. Doesn't seem that hard, AFC 2.0's travails aside.

But the former seems a lot stickier. Consider this highly sketchy sketch of the arrangement of the Porter ticket hall and commuter rail tracks:

View attachment 41738

The CR ROW is too narrow to quad-track things, so you'd need to run the GL trains underneath. But the CR platform is only about 1 story above the ticket hall floor level. It seems like you're left with one of two fairly expensive propositions: A) digging the GL tunnel even deeper, below the floor level of the ticket hall, or b) digging the ticket hall down so its floor level is below the GL tracks. It looks like far more than a simple cut-and-cover job, right? Or does the fact that the GL tracks would effectively be at only a "sub-basement 2" level thanks to the CR tracks already existing in a trench make it less daunting?

View attachment 41739

@F-Line to Dudley's thinking on this a couple of years ago was to put the Green Line tracks immediately under the commuter rail tracks. I haven't looked particularly closely at it myself. As I read it, the Green Line tracks would be accessed via the commuter rail platform (I think).

 

[Charlie_mta]

I love this idea in all ways except one: it makes a future extension beyond Porter much less feasible.

With an indefinite terminus of Porter, though, it works great, in my estimation. No need to fare-gate the Green Line platform, if the other GLX stations are already not fare-gated. You could even build a new headhouse and vertical transportation at Somerville Ave opposite Mossland for better GL and CR access for Spring Hill.

If it were up to me, I would just elevate the GLX extension over Beacon Street and Mass Ave, centered above the existing RL Porter Sq headhouse that spans the Fitchburg Div. The elevated structure alignment would be right above the Fitchburg Div tracks. Its touchdown points would be east of Beacon Street and west of Mass Ave. An elevated line would be the cheapest and least disruptive of any option for a line continuing further west from Porter Sq.

 

[Aprehensive_Words]

@F-Line to Dudley's thinking on this a couple of years ago was to put the Green Line tracks immediately under the commuter rail tracks. I haven't looked particularly closely at it myself. As I read it, the Green Line tracks would be accessed via the commuter rail platform (I think).

I saw that, but is that really workable? That'd force all pax to go up to street level to change between the three lines that would be intersecting there because those GL tunnels would wipe out the lobby-level connection to the Red Line.

Alternately, I suppose you could make the transfer via a Park Street GL northbound-style level crossing for pedestrians, but I'd worry you'd have to set the platform height too low. You don't have a lot of distance between the Red Line ticket hall and the notional Green Line tracks to transition from one floor level to the other at an ADA-compliant slope. I guess the two key questions are: Do the Porter CR tracks need to be lowered at all to add catenary for those trains, and what do the foundations look like for the substation that hangs over the CR tracks? Are those foundations set up to allow you to send a ramp (light blue, below) parallel to the GL tracks down to whatever level you need? And is that ramp wide enough to allow an adequate flow of passengers?

EDIT: This discussion is making me wonder if it would have been better if the T had decided to serve Union with regional rail instead of a branch of the Green Line

 

[Riverside]

I saw that, but is that really workable? That'd force all pax to go up to street level to change between the three lines that would be intersecting there because those GL tunnels would wipe out the lobby-level connection to the Red Line.

Alternately, I suppose you could make the transfer via a Park Street GL northbound-style level crossing for pedestrians, but I'd worry you'd have to set the platform height too low. You don't have a lot of distance between the Red Line ticket hall and the notional Green Line tracks to transition from one floor level to the other at an ADA-compliant slope. I guess the two key questions are: Do the Porter CR tracks need to be lowered at all to add catenary for those trains, and what do the foundations look like for the substation that hangs over the CR tracks? Are those foundations set up to allow you to send a ramp (light blue, below) parallel to the GL tracks down to whatever level you need? And is that ramp wide enough to allow an adequate flow of passengers?

View attachment 41743

EDIT: This discussion is making me wonder if it would have been better if the T had decided to serve Union with regional rail instead of a branch of the Green Line

Yeah, I really don't know about the logistics of building the platform. I haven't been to Porter in a while, but I was able to find diagrams from the original 1977 EIS:

The Green <> Red transfer wouldn't have to be up to street level -- you could add a pedestrian passage under the Green Line tracks that then comes up to the existing fare lobby. (I think.)

But yes, to your original point, it does seem like something will need to go one level below the current lobby -- whether that's a underpassage, the Green Line platforms (with a mezzanine directly above, under the commuter rail platforms), or digging out the current lobby to drop one level down.

A Park St-style level crossing seems dicey. Even if we aren't planning on a Green Line HRT conversion, I still think it makes sense to design with HRT philosophy in mind, for capacity and reliability, etc etc.