Crazy Transit Pitches

Even the most "out there" crazy transit pitch would only add one other potential corridor, along Route 1 (i.e. a heavy rail extension):

View attachment 46153

But for the most part, we're talking about those first four. Whether it's just high-frequency buses or full-out dual LRT lines, Washington and Broadway will both remain the major radial corridors, and the Grand Junction + Eastern Route is always going to be the main corridor running east-west.
Strictly for funsies only: I did propose this alignment for an alternative Urban Ring that runs down Everett Ave into more central areas of Chelsea and then Eastern Ave (which I'm sure you know). It's a consideration for a deep bored tunnel across Chelsea Creek that does not depend on the suspension bridge: if we're spending the $$$ for it, might as well bore it in a way that serves Chelsea better. But otherwise, I agree that it doesn't fundamentally change your analysis.


On the question of space: looking at the (new) current Chelsea station, as well as the new center-platform Boston Landing station, a 2-track mainline station requires about 45 feet of ROW width. The tracks alone require about 30 feet of width. Rapid transit tracks require about 25 feet of width; the GLX stations require 45 feet of width (the GLX + Lowell total width is about 63 feet, by my measurements).
  • Mainline 2 tracks: 30 feet
  • Mainline 2 tracks + platform: 45 feet
  • Rapid transit 2 tracks: 25 feet
  • Rapid transit 2 tracks + platform: 45 feet
If a mainline station and rapid transit station are placed side-by-side, the ROW would probably need to be 90 feet wide. I don't really see any place within the circled area above that could accommodate that.

But, staggered stations might relieve some of the pressure to squeeze full-width stations side by side (75 feet at the rapid transit station, 70 feet at the mainline), something like this:

View attachment 46156

It looks to me that the ROW between Washington and Broadway is just about 70 feet. The ROW to the west widens a bit to 80 feet (actually wider than that if you include the old mainline platform, but its spacing seems like it would be mostly useless for tracks and platforms), though lengthening the current SL3 station to GLX size might be a little more challenging.

I'm not sure if 70-75 ft is enough for a platform and 4 tracks. The four GLX stations alongside Lowell Line all measure almost exactly 80 ft, even in areas tightly constrained on space, like Ball Square:
1703729504990.png


Having said that, there do seem to be plenty of examples of commuter rail platforms that are 10 ft wide. Your maximum for the CR platform will be 10-15 ft here, which is certainly less than ideal, but may be doable for a single island platform.

A minor correction to your diagram is that the 800 ft CR platform needs to span underneath and east of Broadway, as the distance between Washington Ave and Broadway is only 640 ft. But that should be pretty doable as the crossing under Broadway is quite spacious.

Setting aside the silliness of making the brand new Chelsea station obsolete, this raises the larger question about the role of the Regional Rail stations, the Circle Line services, and the necessities of transfers to/from the T111 and T116.

My take is that it makes sense to relocate the mainline station to Washington / Broadway after NSRL is built and mainline trains at Chelsea are through-run to South Station and beyond; at that point, being able to transfer from the T111 and T116 (a la Sullivan Square) becomes more valuable. Until then, mainline passengers who are commuting, e.g., Salem <> Airport can transfer to Circle services at the current Mystic Mall station.

Would frequent non-NSRL Regional Rail services tip the scales in favor of relocation? I still think probably not -- I don't see many T111 or T116 riders transferring to North Station-bound Regional Rail instead of riding to Haymarket or transferring at Maverick.
A few factors to keep in mind:
  • The T111 is the second most frequent bus route in the BNRD (only behind the combined SL4/5), and the T116 is also slightly more frequent than other Frequent "T" Bus Routes, mostly due to off-peak frequencies. (In terms of # trips per weekday, the T116 is only behind SL4/5, T111, T32, T57 and T1 in this order, and tied with T28.) So yes, I do think it's very worthwhile to improve connection from the T111/T116 to both CR and Urban Ring.
  • If you want to promote both T111 and T116 to true BRT (or even street-running LRT), the Washington Ave bridge only has space for 3 travel lanes. This means a very plausible approach will be to set up one transit lane in each direction along Washington Ave, Broadway and Cary Ave, and reroute both T111 and T116 like below. This is a scenario that further increases the need for connections at Broadway (i.e. not just Washington Ave).
T111 T116 BRT Chelsea.png

Another note is that there's a need for better connection from Urban Ring to Broadway, which can be achieved by moving Box District station further west.

In my recent fantasy map the option I went with for serving Chelsea was a full C&C subway under Broadway, which generally does a better job of hitting the major locations in Chelsea compared to the SL3 alignment. It can also be extended further into Revere Center, another major location with extremely high bus ridership that would significantly benefit from better transit.
To be clear, I agree that Broadway from Chelsea to Revere deserves better radial transit, and I even made god-mode proposals for it before. But in reality, I'm not sure if a subway will be built there even with C&C. Broadway is heavily residential, so the drawbacks of C&C - which ironically Jeff just pointed out above - will be especially salient. (Put it another way, the political challenges for a Chelsea-Revere C&C tunnel are almost exactly the same as the 66 corridor from Charles River to Muddy River, and you know my opinions on that.)

In terms of the southern side, the replacement of the Tobin Bridge in the near-ish future means there's a good opportunity to run a rail line across a new bridge
How sure are we about that? I certainly hope so, and IIRC the design guidelines require bus lanes, but I'm not sure if they will necessarily be built in a way that allows (easy) conversion to rail in the future. The grades from I-93 to the river may also be too steep for an intermediate station in Charlestown.

which would also then serve Charlestown, another underserved area, before meeting the OL at Community College and then doing... something, there's options.
This would be a good fit for F-Line's "Red X" proposal. (Which is surprisingly ambitious by F-Line standards.)

Maybe this could be combined with another C&C subway under Broadway (The Everett one) that branches off the OL?
Any OL branch that cuts the frequencies of Malden, Wellington and Assembly permanently in half is a non-starter, not to mention this proposal even hurts Sullivan. The real-world MBTA Orange Line Transformation aims for 3-min headways, which demonstrates a real need for that coming from Malden, and cutting that to 6 min will make you even worse off than pre-Covid levels. Having 3-min headways on both branches require 1.5-min headways downtown, which only the absolute top tier heavy rail systems worldwide can achieve.

(Edit: Branching OL was discussed here, but I stand by my original opinion. All the OL Transformation work shouldn't result in a downgrade to 6-min frequency with no way of improving further.)

I went far down the crayoning rabbit hole here, well outside the bounds of feasibility, but useful as a point of imagination:

View attachment 46161
To clarify, are your Bronze Line Everett/Revere branches grade separated in this scenario? If they're not (which itself is reasonable given the Red X), then I don't see the two Silver Line branches being grade separated either, especially the Malden branch.
 
Last edited:
I swear it’s purely coincidence I mentioned the EL on Rt 1 in the other thread and you’re showing a line along the same route here ;). I’ve always imagined your red line diverting from Rt 1 briefly along Everett Ave to a large, consolidated Chelsea Sta. at Mystic Mall. I think that‘s prime land for a sort of “New Downtown” for Chelsea, and you can maintain a good connection the current downtown with a station at the Rt 1/Everett Ave interchange
Hahaha, I think it is indeed coincidence and that I only saw your post afterward.

That is an interesting idea about diverting down Everett Ave. A T111 connection would still be possible at Everett Ave/Rt 1, but the T116 is harder to swing (which certainly isn't the end of the world, but is a downside).
I'm not sure if 70-75 ft is enough for a platform and 4 tracks. The four GLX stations alongside Lowell Line all measure almost exactly 80 ft, even in areas tightly constrained on space, like Ball Square:
....
Having said that, there do seem to be plenty of examples of commuter rail platforms that are 10 ft wide. Your maximum for the CR platform will be 10-15 ft here, which is certainly less than ideal, but may be doable for a single island platform.
Hmm, yeah I need to look at that again. The impression I get is that the new standard is for platforms to be at least 15 feet wide. So that may well be the fatal flaw here.
If you want to promote both T111 and T116 to true BRT (or even street-running LRT), the Washington Ave bridge only has space for 3 travel lanes. This means a very plausible approach will be to set up one transit lane in each direction along Washington Ave, Broadway and Cary Ave, and reroute both T111 and T116 like below. This is a scenario that further increases the need for connections at Broadway (i.e. not just Washington Ave).
Yes, great point here.
Another note is that there's a need for better connection from Urban Ring to Broadway, which can be achieved by moving Box District station further west.
It seems like this is what we are landing on. That said, the walk down to Box District station from Broadway is actually pretty short, so a relocation might not be super urgent.
To clarify, are your Bronze Line Everett/Revere branches grade separated in this scenario? If they're not (which itself is reasonable given the Red X), then I don't see the two Silver Line branches being grade separated either, especially the Malden branch.
Honestly I didn't even think about it to that level of detail. Yes, in any slightly realistic version of this, both Silver and Bronze are surface-running. But, that in turn makes it hard to run them to the full length sketched out here -- probably would be slightly more realistic with the western branch truncated to Glendale and the eastern branch truncated to... idk, Park Ave or Chelsea Common?
 
Any OL branch that cuts the frequencies of Malden, Wellington and Assembly permanently in half is a non-starter, not to mention this proposal even hurts Sullivan. The real-world MBTA Orange Line Transformation aims for 3-min headways, which demonstrates a real need for that coming from Malden, and cutting that to 6 min will make you even worse off than pre-Covid levels. Having 3-min headways on both branches require 1.5-min headways downtown, which only the absolute top tier heavy rail systems worldwide can achieve.

(Edit: Branching OL was discussed here, but I stand by my original opinion. All the OL Transformation work shouldn't result in a downgrade to 6-min frequency with no way of improving further.)
Do you have full historical frequency information? The best I can find is an article from 2013 which says peak frequencies were 6 minutes. And so the next question is, does the northern end of the line actually need higher frequencies, or would the main benefit be felt downtown? I think with a 2nd OL branch to Everett, which would likely take a significant amount of the bus transfers away from Wellington and Malden Center, it would be the latter. I also don't think we should necessarily be too concerned with the limitations of the signal system now (Or in a few years at least) when realistically anything talked about here is likely 15 years out minimum. There's time to make some upgrades.

That's not to say that Malden never should get higher frequencies, but I don't think it would be an immediate need in an OL to Everett world, and with CBTC you could absolutely push frequencies higher in the further future.
 
Do you have full historical frequency information? The best I can find is an article from 2013 which says peak frequencies were 6 minutes. And so the next question is, does the northern end of the line actually need higher frequencies, or would the main benefit be felt downtown? I think with a 2nd OL branch to Everett, which would likely take a significant amount of the bus transfers away from Wellington and Malden Center, it would be the latter. I also don't think we should necessarily be too concerned with the limitations of the signal system now (Or in a few years at least) when realistically anything talked about here is likely 15 years out minimum. There's time to make some upgrades.

That's not to say that Malden never should get higher frequencies, but I don't think it would be an immediate need in an OL to Everett world, and with CBTC you could absolutely push frequencies higher in the further future.
For pre-Covid headways, the historical system maps (thanks to Delvin) shows 6-min headways during rush hours in winter 2020. However, IIRC there were already overcrowding on both Red and Orange Lines, hence the need for the Transformation programs.

For existing ridership and whether 6-min headways are enough, this figure summarizes it all:
OL load Fall19.png

While the absolute peak load occurs at North Station, the load (which determines capacity needed) gets very close to peak at Sullivan (93% of North Station) and even at Assembly (80%). North Station's AM peak boardings are only 50% of Malden Center's, 22% of Oak Grove to Wellington combined, and 18% of Oak Grove to Sullivan combined; not to mention a good chunk alight at North Station. NSRL will remove North Station's boardings even further. So from a demand perspective, the only feasible branching point is at least Wellington or preferably Malden Center.

For bus ridership from OL to Sullivan, TL;DR: It doesn't make a difference to the load chart above. The # of Everett bus riders at the three bus hubs is tiny compared to # of all bus riders, which itself is small compared to # of all OL riders.

I'm reusing the chart for Sullivan that I made before:
Sullivan bus riderships Fall22.png

The 104 and 109, which serve Everett, do contribute to 40% of Sullivan's bus ridership, and they're likely to be diverted away with a radial Everett branch. (Including 92 and 93 raises this to 48%.) However, we've already seen from above that Sullivan's AM peak boardings is behind each of Oak Grove, Malden Center and Wellington.

A very quick replication for Wellington:
Wellington bus riderships Fall22.png

The 110 is the most major single bus route, though it only contributes 29% of Wellington's ridership (lower share than 104/109 at Sullivan). It's also important to note that ridership on the 110 is not just limited to the Everett Broadway corridor: its bus route profile shows even distribution all the way to Route 1.
  • On outbound 110, 675.9 weekly riders (23% of Wellington's boardings) alight on or before Chelsea St in the Everett Square vicinity.
  • If you consider Ferry St to be within the walkshed of a Glendale Square station, that amounts to 1822.6 weekly riders (62% of Wellington). Some of these may have boarded at Everett instead of Wellington, as well.
In the grand scheme of things, an Everett Broadway subway is unlikely to siphon off significant bus ridership from Wellington.

Last but not least, Malden Center:
Malden Center bus riderships Fall22.png

The 104 does overlap an Everett Broadway subway, but it's barely the top bus route at Malden (19% of all). Most buses don't have significant overlaps (that includes 99 and 106, whose Main St corridor is not very close to Broadway). Even inbound 104 has equal ons and offs at Ferry St @ Broadway, and switches to boarding mode after that - most Malden 104 riders alight at Ferry St. Assuming a Glendale Square station's walkshed spans from Bennett St to Maple Ave, these stops have 1624.3 weekly inbound alightings.

Now I do a coarse estimate of the total impacts on peak load from removing Everett Broadway riders. The 2015-17 passenger survey data indicate 66%, 58% and 53% of riders boarding at Sullivan, Wellington and Malden Center connect from buses respectively. So:
StationFall19 Load (AM peak)# Bus Riders*# Everett Bus Riders**Load without Everett Bus Riders
Oak Grove31893189
Malden Center767124412587413
Wellington1024915331719820
Assembly1075310324
Sullivan Square12413125850611478
Community College1237611441
North Station1339012455

* Multiplies Fall19 AM peak inbound boardings by % of bus transfers from 2015-17 passenger survey data
* Multiplies # bus riders by the ridership proportions of the following routes in Fall22 data: (Sullivan) 100% of 104 and 109; (Wellington) 100% of 110 up to Chelsea St, and 40% of 110 on Ferry St; (Malden) 100% of 104 in the Glendale walkshed


I'll just repeat the TL;DR: It doesn't make a difference to the load chart above. The # of Everett bus riders at the three bus hubs is tiny compared to # of all bus riders, which itself is small compared to # of all OL riders.
 
Last edited:
For pre-Covid headways, the historical system maps (thanks to Delvin) shows 6-min headways during rush hours in winter 2020. However, IIRC there were already overcrowding on both Red and Orange Lines, hence the need for the Transformation programs.

For existing ridership and whether 6-min headways are enough, this figure summarizes it all:
View attachment 46165
While the absolute peak load occurs at North Station, the load (which determines capacity needed) gets very close to peak at Sullivan (93% of North Station) and even at Assembly (80%). North Station's AM peak boardings are only 50% of Malden Center's, 22% of Oak Grove to Wellington combined, and 18% of Oak Grove to Sullivan combined; not to mention a good chunk alight at North Station. NSRL will remove North Station's boardings even further. So from a demand perspective, the only feasible branching point is at least Wellington or preferably Malden Center.

For bus ridership from OL to Sullivan, TL;DR: It doesn't make a difference to the load chart above. The # of Everett bus riders at the three bus hubs is tiny compared to # of all bus riders, which itself is small compared to # of all OL riders.

I'm reusing the chart for Sullivan that I made before:
View attachment 46166
The 104 and 109, which serve Everett, do contribute to 40% of Sullivan's bus ridership, and they're likely to be diverted away with a radial Everett branch. (Including 92 and 93 raises this to 48%.) However, we've already seen from above that Sullivan's AM peak boardings is behind each of Oak Grove, Malden Center and Wellington.

A very quick replication for Wellington:
View attachment 46168
The 110 is the most major single bus route, though it only contributes 29% of Wellington's ridership (lower share than 104/109 at Sullivan). It's also important to note that ridership on the 110 is not just limited to the Everett Broadway corridor: its bus route profile shows even distribution all the way to Route 1.
  • On outbound 110, 675.9 weekly riders (23% of Wellington's boardings) alight on or before Chelsea St in the Everett Square vicinity.
  • If you consider Ferry St to be within the walkshed of a Glendale Square station, that amounts to 1822.6 weekly riders (62% of Wellington). Some of these may have boarded at Everett instead of Wellington, as well.
In the grand scheme of things, an Everett Broadway subway is unlikely to siphon off significant bus ridership from Wellington.

Last but not least, Malden Center:
View attachment 46169
The 104 does overlap an Everett Broadway subway, but it's barely the top bus route at Malden (19% of all). Most buses don't have significant overlaps (that includes 99 and 106, whose Main St corridor is not very close to Broadway). Even inbound 104 has equal ons and offs at Ferry St @ Broadway, and switches to boarding mode after that - most Malden 104 riders alight at Ferry St. Assuming a Glendale Square station's walkshed spans from Bennett St to Maple Ave, these stops have 1624.3 weekly inbound alightings.

Now I do a coarse estimate of the total impacts on peak load from removing Everett Broadway riders. The 2015-17 passenger survey data indicate 66%, 58% and 53% of riders boarding at Sullivan, Wellington and Malden Center connect from buses respectively. So:
StationFall19 Load (AM peak)# Bus Riders*# Everett Bus Riders**Load without Everett Bus Riders
Oak Grove31893189
Malden Center767124412587413
Wellington1024915331719820
Assembly1075310324
Sullivan Square12413125850611478
Community College1237611441
North Station1339012455

* Multiplies Fall19 AM peak inbound boardings by % of bus transfers from 2015-17 passenger survey data
* Multiplies # bus riders by the ridership proportions of the following routes in Fall22 data: (Sullivan) 100% of 104 and 109; (Wellington) 100% of 110 up to Chelsea St, and 40% of 110 on Ferry St; (Malden) 100% of 104 in the Glendale walkshed


I'll just repeat the TL;DR: It doesn't make a difference to the load chart above. The # of Everett bus riders at the three bus hubs is tiny compared to # of all bus riders, which itself is small compared to # of all OL riders.
Excellent analysis. Clearly I overestimated the impact those buses have on OL ridership. However, I'm still not sure the load data suggests huge headway improvements are necessary north of Sullivan. Disregarding any bus ridership difference, and using the passenger capacity of the old trains since I can't immediately find capacities for the new trains, and as you're about to see it doesn't end up mattering that much. (131 regular, 224 max per car, 6 cars for 786 and 1344 pax respectively.) With the 2019 load numbers, assuming 6 minute headways for a total of 20 trains between 7AM and 9AM, each train (on average) heading south from Assembly would have 537 passengers. A bit crowded, but more than enough for everyone to have at least some personal space. Even if we double the load, there would still be enough room, although it would start to become a bit uncomfortable. The new trains almost certainly improve the situation somewhat, along with being able to actually meet frequency targets when your trains aren't 50 year old rust buckets.
 
Improving service on Warren St
This is an interesting and very relevant thought, which I've also briefly considered before. Indeed, if we use Delvin's map as a rough measure of density:
View attachment 45047
While there's good residential density (and job density?) around Nubian, Warren St is arguably even denser. And it nicely fills the gap between Orange and Fairmount, effectively covering most of the area with their walksheds. Even nicer if we can run it to Franklin Park (a recreational destination itself) and/or connect to Fairmount at Four Corners/Geneva.
I'm late to the party here and probably won't be able to respond to everything in the subsequent discussion, but I'll try to layer in thoughts where I can.
Another issue is that while the northern half of Warren St is easily wide enough, the southern half between Quincy St and Blue Hill Ave is narrow and likely cannot not support bi-directional transit lanes. A few solutions:
  • Terminate the line at MLK Blvd (not ideal for catchment)
I agree that an MLK terminal is far from perfect, but I would not quite so quickly dismiss its value. (Letting the perfect become the enemy of the good.) MLK is also useful because it's wide enough that you could have an on-street terminal where trains can turn without blocking the (presumably shared bus/LRT) transit lanes on Warren.
  • Dig a short, 0.5-mile tunnel that bypasses this narrow section of Warren St ($$)
I can check this again, but my recollection from the last time I looked at this, the problem was actually about where to put the portals, especially at the southern end.
  • Run in mixed traffic (lower reliability, not sure if traffic is decent enough)
I think this depends whether your goal is an OSR to downtown, or something else. If the former, then I think mixed traffic isn't ideal.
  • Or, this interesting idea that takes 1 lane each on Warren St, BHA and Quincy St: (some deviations from typical route, and may either still run in mixed traffic or turn these streets into one-way)
View attachment 45048
Yeah, this is an interesting thought. Yeah, would probably need to be one-way, or obliterate lots of parking -- if you're going to run in mixed traffic, why deviate from Warren at all?

Perhaps this could be somewhat useful post-BNRD to improve reliability on the bus routes? On the other hand, losing northbound service on Warren in favor of a 4-min walk to Blue Hill doesn't sound amazing.
(If we build all of this, Washington streetcar and I-93 subway, I think it's worth looking at letting the I-93 service serve the surface section south of Nubian rather than the streetcar. In addition to avoiding a B branch 2.0, this also improves travel times for points south of Nubian, and lets you dig a tunnel through Nubian onto a portal on Warren St that bypasses several big intersections around Nubian. This point may be more for @Riverside.)
Yeah, this of course goes back to my analysis of the strangeness of the B and C Lines. I think there are pros and cons. On the one hand, Warren-via-subway would provide a faster OSR to downtown. On the other hand, potentially you degrade the reliability of the subway portion, which we hope will be used by bus riders from all across Dorchester.

To me, from a network design perspective, the "natural" portal point for a subway would actually be Grove Hall. That's where the network blooms outward in different directions:

1703789478413.png


Depending how far you go down Blue Hill, that gives you surface segments between 1.5 and 3.3 miles long, which seems much more reasonable.

Of course, it'll be hard enough to build a subway (or viaduct) to Nubian. Doubling it (and along an alignment that I think would need a TBM) makes it all the more difficult.

Anyway, I've wandered far afield from your original question. I still feel that the best option all around to enhance transit access to the Warren St corridor is an extension of the Washington surface line down to MLK, combined with whatever the next generation of the T28 etc will be. It's far from perfect, but it provides benefit without particularly harming other parts of the system, and would not require extreme cost.

The other thing to consider is that there are two reasons to have an OSR downtown: to get to downtown, and to transfer to the other lines. But that second purpose can be achieved without going downtown. A line running out of Longwood would have transfers to Green and Orange at the north, and Red (somewhat less ideally) to the south. If extended further to Kenmore, with BLX, you've got all of them. (I guess except for SL3.)

(A separate line running down Columbia Road from JFK/UMass could enhance the above by providing a Red transfer that doesn't require doubling back.)
I'm not sure 5 min headways are actually unworkable. Some quick digging turned up this paper which mentions bunching on the Silver Line, pointing out that the biggest problems for reliability are:
  • Long boarding times, exacerbated at the time of study by the new fare system. This could be improved with off-board fare collection and by using vehicles with more doors. You can even do this with buses, 4 door articulated buses are a thing that exist. (Although these are slightly longer)
  • Because Temple Place is constrained, it doesn't really act like a terminus. The line therefore operates like it's twice as long, with buses having to correct for deviations in headways from both the inbound and outbound trips at Nubian. This would be alleviated by having a proper downtown terminus.
  • The biggest factor according to the study, regulation of headways at Nubian is (Or at least was)... crap, to put it bluntly. Not necessarily because it needs to be, but just from poor operations and operations management. The solution here is pretty self-explanatory.

    With all of these problems addressed, I don't think 5 min headways are impossible.
My problem is that higher headways become less compatible with transit priority signaling. @The EGE has much more detailed analysis on this, but it basically boils down to: the more frequent your transit vehicles come, the harder it is to guarantee that they always get the priority at the intersection. Even at simple intersections, you have through traffic, cross traffic, and pedestrian traffic, and you can't make any group wait too long because it'll back up into other intersections.

So, yes, perhaps 5 minute headways are possible, but it may come at the expense of running time and speed. (And to some extent, reliability.)

(That's a really cool paper, though, thanks for finding it!)

Not sure I have much to add on the other topics that were addressed; I agree with @Teban54 that capacity at Park St is not a concern, and there's definitely no need to turn 3 branches at Park St.

Nothing in particular to add about an LMA subway at the moment. Definitely think you have an interesting analysis about circumferential corridors, @TheRatmeister!
 
I'll just repeat the TL;DR: It doesn't make a difference to the load chart above. The # of Everett bus riders at the three bus hubs is tiny compared to # of all bus riders, which itself is small compared to # of all OL riders.
Excellent analysis. Clearly I overestimated the impact those buses have on OL ridership. However, I'm still not sure the load data suggests huge headway improvements are necessary north of Sullivan. Disregarding any bus ridership difference, and using the passenger capacity of the old trains since I can't immediately find capacities for the new trains, and as you're about to see it doesn't end up mattering that much. (131 regular, 224 max per car, 6 cars for 786 and 1344 pax respectively.) With the 2019 load numbers, assuming 6 minute headways for a total of 20 trains between 7AM and 9AM, each train (on average) heading south from Assembly would have 537 passengers. A bit crowded, but more than enough for everyone to have at least some personal space. Even if we double the load, there would still be enough room, although it would start to become a bit uncomfortable. The new trains almost certainly improve the situation somewhat, along with being able to actually meet frequency targets when your trains aren't 50 year old rust buckets.
Yeah, I second the praise on the analysis. I started to do something similar but got discouraged and then got pulled away.

I do think there is a way to have our cake and eat it too, which is to go back to BERy's original plans for what became the Orange Line: extend it from Everett to Malden.

1703793124310.png


Two (or three) options for doing this:
  • Ferry St: most indirect but follows the strong Broadway corridor for the longest
  • Main St: more direct, but demand is lower; does allow you to serve some major employers in Everett, though it avoids Everett's downtown
  • Saugus Branch: didn't even map this one; the least invasive, but also doesn't really go where people live, and it doesn't even do that well serving Everett's non-downtown employers (compared to the Main St alignment)
But the problems of the Ferry St alignment are more or less the same as they would be for a straight-up branch line like you are proposing, so no big difference there.

Diverting north of Glendale shouldn't be a huge problem. Density north of Glendale is significantly lower, and demand north of Glendale is significantly lower:

1703792475763.png


And Everett and Malden have "islands" of employment centers (as opposed to a steady corridor), meaning it's less important how you connect the two of them (i.e. via Main St vs Ferry St):

1703792625487.png


Finally -- and this is really just a bit of crayoning fun -- you can still use the Western Route alignment for rapid transit. For example, an LRT branch from the Central Subway via Lechmere + Sullivan could take it over, and then swing east at Malden to serve the (surprisingly high demand) Salem St corridor via the Saugus Branch (which could be done at-grade with LRT, as opposed to the full grade sep needed for HRT). Meanwhile, the Orange Line could continue on to Oak Grove and beyond.

~~~

Either way, I think a fully grade separated service to Everett is both a very good idea and an extremely tall order. Whether going to Linden (as you proposed) or to Malden Center, we're still talking about 2-3 miles of elevated or cut-and-cover or TBM subway through the main arterials of the region, along ROWs that have never had grade separated rail. Other than the Red Line's Alewife extension, I don't think we've done something like that in over 100 years.

So, I think it's very hard. BUT, if we are assuming the ~$2B-$4B can be found to do a branch to Linden, it should be equally possible to divert back to Malden, which solves the "max freqs to bus hubs" problem.
 
Yeah, I second the praise on the analysis. I started to do something similar but got discouraged and then got pulled away.

I do think there is a way to have our cake and eat it too, which is to go back to BERy's original plans for what became the Orange Line: extend it from Everett to Malden.

View attachment 46175

Two (or three) options for doing this:
  • Ferry St: most indirect but follows the strong Broadway corridor for the longest
  • Main St: more direct, but demand is lower; does allow you to serve some major employers in Everett, though it avoids Everett's downtown
  • Saugus Branch: didn't even map this one; the least invasive, but also doesn't really go where people live, and it doesn't even do that well serving Everett's non-downtown employers (compared to the Main St alignment)
But the problems of the Ferry St alignment are more or less the same as they would be for a straight-up branch line like you are proposing, so no big difference there.

Diverting north of Glendale shouldn't be a huge problem. Density north of Glendale is significantly lower, and demand north of Glendale is significantly lower:

View attachment 46173

And Everett and Malden have "islands" of employment centers (as opposed to a steady corridor), meaning it's less important how you connect the two of them (i.e. via Main St vs Ferry St):

View attachment 46174

Finally -- and this is really just a bit of crayoning fun -- you can still use the Western Route alignment for rapid transit. For example, an LRT branch from the Central Subway via Lechmere + Sullivan could take it over, and then swing east at Malden to serve the (surprisingly high demand) Salem St corridor via the Saugus Branch (which could be done at-grade with LRT, as opposed to the full grade sep needed for HRT). Meanwhile, the Orange Line could continue on to Oak Grove and beyond.

~~~

Either way, I think a fully grade separated service to Everett is both a very good idea and an extremely tall order. Whether going to Linden (as you proposed) or to Malden Center, we're still talking about 2-3 miles of elevated or cut-and-cover or TBM subway through the main arterials of the region, along ROWs that have never had grade separated rail. Other than the Red Line's Alewife extension, I don't think we've done something like that in over 100 years.

So, I think it's very hard. BUT, if we are assuming the ~$2B-$4B can be found to do a branch to Linden, it should be equally possible to divert back to Malden, which solves the "max freqs to bus hubs" problem.
Very quick comment on this (I'll get to the other points later).

A big issue with diverting OL to Everett (while a very innovative idea itself) is that it skips Assembly, which is a textbook example of successful TOD with residential, commercial and recreational demands. If we're compensating Assembly with another rapid transit line anyway, it would imply two lines, at which point it might make more sense to just keep OL untouched and send the other line into Everett and Glendale (then either swing back to Malden via the T104 or follow the T109 to Linden).

Replacing Wellington's successful P&R is another concern, but a bit less so.

In comparison, I'm less concerned with being close to downtown Everett, especially if cost is a concern. Malden Center OL station is actually not very close to the true "Malden Center", and takes a 9-min walk:
1703802203747.png


On the other hand, here are the walking distances from downtown Everett to:
  • Main St: 7 min
  • Saugus Branch: 10 min
  • Sweetser Circle: 10 min
 
Last edited:
we're still talking about 2-3 miles of elevated or cut-and-cover or TBM subway through the main arterials of the region, along ROWs that have never had grade separated rail. Other than the Red Line's Alewife extension, I don't think we've done something like that in over 100 years.
Well, that's the elephant in the room for 90% of Crazy Transit Pitches nowadays. Most ideas involving tunnels are simply very unlikely to happen in practice - and the question really becomes how much you're willing to assume.

(I'm willing to make an exception for GL Reconfiguration, but its C&C will have much lower financial and social cost compared to corridors like any streets in Everett, Revere, Allston and more, which require both mapping and relocation of utilities and fencing off people's driveways for years.)
 
I randomly think of this idea but never posted the idea cause im not sure if someone else already did. If money wasn't an issue could we technically create a little bend on the Red Line between Broadway and South Station so that we could create an infill stop between them that gets riders closer to the Seaport Area than South Srarion?
 
I randomly think of this idea but never posted the idea cause im not sure if someone else already did. If money wasn't an issue could we technically create a little bend on the Red Line between Broadway and South Station so that we could create an infill stop between them that gets riders closer to the Seaport Area than South Srarion?
This is very thought-provoking. My best effort gave something like this:
1703842665873.png


Moving Broadway two blocks east may actually be beneficial, as the new station is closer to density and reduces awkward routings of the 9 and 11 buses.

The obvious problem is cost. This requires a new crossing of the Fort Point Channel under Summer St (the current RL tunnel starts the turn before hitting water), and crossing under the Big Dig in Seaport with widely spaced tunnels. Perhaps such a reroute is more justifiable in a Red X world, where one of the current RL branches takes this route (connecting to Old Colony on one end and Congress St on the other), and the other branch continues on the current RL route?
 
Excellent analysis. Clearly I overestimated the impact those buses have on OL ridership. However, I'm still not sure the load data suggests huge headway improvements are necessary north of Sullivan. Disregarding any bus ridership difference, and using the passenger capacity of the old trains since I can't immediately find capacities for the new trains, and as you're about to see it doesn't end up mattering that much. (131 regular, 224 max per car, 6 cars for 786 and 1344 pax respectively.) With the 2019 load numbers, assuming 6 minute headways for a total of 20 trains between 7AM and 9AM, each train (on average) heading south from Assembly would have 537 passengers. A bit crowded, but more than enough for everyone to have at least some personal space. Even if we double the load, there would still be enough room, although it would start to become a bit uncomfortable. The new trains almost certainly improve the situation somewhat, along with being able to actually meet frequency targets when your trains aren't 50 year old rust buckets.
Here is a comment saying that pre-Covid, OL gets absolutely packed at Community College to the point that people couldn't get on. That doesn't sound like "at least some personal space" to me. Not sure if it's a result of the load data being underestimates, or uneven averages during the 7-9am time frame.

The new trains and a general reduction of ridership post-Covid may help with things, but I don't think they change the calculus by much, especially if future growth and extension to Reading are considered.
 
Well, that's the elephant in the room for 90% of Crazy Transit Pitches nowadays. Most ideas involving tunnels are simply very unlikely to happen in practice - and the question really becomes how much you're willing to assume.

(I'm willing to make an exception for GL Reconfiguration, but its C&C will have much lower financial and social cost compared to corridors like any streets in Everett, Revere, Allston and more, which require both mapping and relocation of utilities and fencing off people's driveways for years.)
It's so tragic that that's true. You look at places like Austria, Korea, Italy, Spain....In Korea in particular, they're tunneling everything, even things they arguably shouldn't like trams. It's not like the U.S.' geology is so special we can't do it; we just have so many other..."concerns"...that we end up not doing even things that are blindingly obvious. But then, we don't even make good use of the assets we already have, I guess it shouldn't raise an eyebrow that we don't do anything new, either.
 
Here is a comment saying that pre-Covid, OL gets absolutely packed at Community College to the point that people couldn't get on. That doesn't sound like "at least some personal space" to me. Not sure if it's a result of the load data being underestimates, or uneven averages during the 7-9am time frame.

The new trains and a general reduction of ridership post-Covid may help with things, but I don't think they change the calculus by much, especially if future growth and extension to Reading are considered.
I suspect there's several things going on here:
  1. Averages are not even during rush hour, and some trains are more busy than others
  2. Camera-based APC systems seem to struggle with crowds, potentially leading to some undercounting
  3. Inconsiderate people with backpacks effectively lower the capacity of trains
  4. The trains being poorly maintained old rust buckets leading to reduced and highly variable headways and more crowding (I'm no expert, but whenever I rode the OL pre-covid it felt like I always had to wait ages and the platforms at stations like DTX were always crowded, even off-peak.)
New trains definitely help with point #4, and they're higher capacity which obviously helps more. Unfortunately this is one of those questions where coming up with a perfect answer just isn't really possible with the available data. It seems like there's some spare capacity for branching north of Sullivan, but maybe future growth would explode and it would be a problem, but I don't think that's the most likely outcome. Even if it did, there would still be things like the introduction of CBTC and automation or platform lengthening to boost capacity if necessary. Given the potential amount of ridership from Everett I think that's a worthwhile tradeoff.
 
If we're compensating Assembly with another rapid transit line anyway, it would imply two lines, at which point it might make more sense to just keep OL untouched and send the other line into Everett and Glendale (then either swing back to Malden via the T104 or follow the T109 to Linden).
Oh yeah, that's a good point. Yeah, I mean, I've thought for a long time that an LRT station at Sullivan could be a jumping off point for rapid transit into Everett. Could be Green/Gold Line from the Central Subway, or "Bronze Line" service from Kendall -- pros and cons to each. Somehow I think I've never considered a radial branch to Malden Center via the Saugus Branch (though IIRC @vanshnookenraggen crayoned one, like, 15 years ago).

Your point about distance to downtown is an interesting one though. Sweetster Circle would need a fair amount of rework anyway if there were a station there, which could perhaps improve pedestrian access. (Or if you could build a good bus transfer station there, that could be interesting too.) That said, I don't see an easy way to place a station on the Saugus Branch; maybe something at Tileston St, but that's in a quiet residential neighborhood and still seems like it's some distance from the action on Broadway.

I suppose you could run a radial line (Gold below) from the Central Subway, out the Saugus Branch, serve the Salem St corridor plus West Everett, and then run the line from Kendall in dedicated lanes along Broadway through the heart of Everett.

1703872180634.png


I've been ambitious here and drawn the Bronze Line all the way to Glendale, but you could probably still see benefit from more modest extensions to Hancock St or Chelsea St. Especially if shorter, that surface-running segment could be controllable, and probably wouldn't be any more disruptive than a Nubian Line along Washington or an extension to Hyde Square.
 
This is very thought-provoking. My best effort gave something like this:
View attachment 46185

Moving Broadway two blocks east may actually be beneficial, as the new station is closer to density and reduces awkward routings of the 9 and 11 buses.

The obvious problem is cost. This requires a new crossing of the Fort Point Channel under Summer St (the current RL tunnel starts the turn before hitting water), and crossing under the Big Dig in Seaport with widely spaced tunnels. Perhaps such a reroute is more justifiable in a Red X world, where one of the current RL branches takes this route (connecting to Old Colony on one end and Congress St on the other), and the other branch continues on the current RL route?
I never posted this but a while ago I crayoned something similar, integrated into the Red X:

1703872984032.png


Now... the interesting thing is that, IIRC, the Silver Line tunnel was actually built to Red Line specifications. So... if you really wanted to blow up the South Station subway station... and you wanted to figure out how to do a stupid 180 degree turn after WTC under all the Mass Pike spaghetti... I think you could do it.

But yeah -- if the Red Line were built now, the bend under Fort Point Channel would be somewhere in the Seaport. It makes sense where it was built -- at the time, that was the edge of Downtown. But now Downtown extends much farther, so...
 
Okay, so... I swear this isn't a shitpost (though it definitely is a hot take).

Crazy Transit Pitch: a cable car from Chelsea's Eastern Ave to the Logan Terminals:

1704047839468.png


@Teban54 and @737900er pointed out that Eastern Ave could be a useful place to add parking for Logan, and could help get Massport to pay for a crossing of Chelsea Creek (somewhat expensive, due to the need for large ships to be able to pass through).

The key question is whether a cable car could be a cheaper way to cross Chelsea Creek (versus building a much higher bridge for rail/bus). I don't know much about cable cars at all, but here's what I've found Googling:

Paris is building a cable car, estimated at about $57M per mile. Portland built a shorter cable car (aerial tram) that apparently had lots of cost overruns in part due to re-planning, which came out to $91.2M per mile. Both of these are lower than the $100M-$275M range I found for (non-GLX) modern light rail projects.

Can a cable car get over Chelsea Creek? Not sure. I couldn't find a lot on maximum gradients for cable cars, but the one in Palm Springs has a gradient of 42 degrees at one point, which is a 90% grade. By my reckoning, from the (current) Eastern Ave station to the shore of the Creek is 537 feet, which at a 90% grade would lift you to a height of about 480 feet.

Do I know whether 480 feet is high enough to let those ships underneath? No I do not.

The other issue is that, for this to be worthwhile, I'd argue that, whatever vehicle you use to cross Chelsea Creek, that same vehicle should take you all the way to the terminals (or at least to Central Parking). So... this would need to be long-ish. (The route above is just over 2.5 miles.)

The good news is that there is public right-of-way pretty much the entire route, which (hopefully) would simplify construction.

EDIT: you’d need a height of something like 175 feet (probably a bit more for cushion, we could perhaps round to ~200).
 
Last edited:
Okay, so... I swear this isn't a shitpost (though it definitely is a hot take).

Crazy Transit Pitch: a cable car from Chelsea's Eastern Ave to the Logan Terminals:

View attachment 46248

@Teban54 and @737900er pointed out that Eastern Ave could be a useful place to add parking for Logan, and could help get Massport to pay for a crossing of Chelsea Creek (somewhat expensive, due to the need for large ships to be able to pass through).

The key question is whether a cable car could be a cheaper way to cross Chelsea Creek (versus building a much higher bridge for rail/bus). I don't know much about cable cars at all, but here's what I've found Googling:

Paris is building a cable car, estimated at about $57M per mile. Portland built a shorter cable car (aerial tram) that apparently had lots of cost overruns in part due to re-planning, which came out to $91.2M per mile. Both of these are lower than the $100M-$275M range I found for (non-GLX) modern light rail projects.

Can a cable car get over Chelsea Creek? Not sure. I couldn't find a lot on maximum gradients for cable cars, but the one in Palm Springs has a gradient of 42 degrees at one point, which is a 90% grade. By my reckoning, from the (current) Eastern Ave station to the shore of the Creek is 537 feet, which at a 90% grade would lift you to a height of about 480 feet.

Do I know whether 480 feet is high enough to let those ships underneath? No I do not.

The other issue is that, for this to be worthwhile, I'd argue that, whatever vehicle you use to cross Chelsea Creek, that same vehicle should take you all the way to the terminals (or at least to Central Parking). So... this would need to be long-ish. (The route above is just over 2.5 miles.)

The good news is that there is public right-of-way pretty much the entire route, which (hopefully) would simplify construction.
This actually doesn't seem that insane. I don't think going Parking-All Terminals is the right move, but I think Central Parking + maybe Terminal B as well since that's quite a walk would be feasible. The only major concern I have is that now we have a big, tall thing next to an airport, which could obviously lead to some problems. I'm not totally sure how high the no-build areas around the airport are so maybe it's fine, or maybe you could work around it? Another potential problem would be accessibility. A lot of "Commuter" cable car lines like what you see in Latin America have boarding where the cars never actually stop. For people with lots of luggage or accessibility limitations this could be a deal-breaker.
 
Just as a cable-car related aside, I have been thinking about where one might be practical. It feels like there should be some route possible that makes a useful transport link while also feeding into the more touristy appeal of a cable car, like some combination of Jeffries Point, Seaport, Long Wharf, and Charlestown Navy Yard. Something over the Charles might also be possible but I don't think that would really be useful, more just neat. (I'm also not sure if the airspace above the Charles is heavily used for helicopters/light aircraft like the Hudson or Thames are in NY/London respectively.)
 
I don't think going Parking-All Terminals is the right move, but I think Central Parking + maybe Terminal B as well since that's quite a walk would be feasible.
Yeah I waffled on this for a while, and then decided, "Well, this is gonna be a problem for anything we build to Logan, so I'll just do this as a placeholder". But yeah, Central Parking + Terminal B was also something I considered.
he only major concern I have is that now we have a big, tall thing next to an airport, which could obviously lead to some problems. I'm not totally sure how high the no-build areas around the airport are so maybe it's fine, or maybe you could work around it?
I was thinking about this -- I figure once you get across the Creek, there's no real need for the cable car to be any higher than a conventional el. And, like the question of which terminal(s) to serve, that'll be a problem for whatever we end up building to Logan.
A lot of "Commuter" cable car lines like what you see in Latin America have boarding where the cars never actually stop. For people with lots of luggage or accessibility limitations this could be a deal-breaker.
Oh that's super interesting, I didn't know that. Yes, these would have to be able to come to a full stop, that's for sure.
 

Back
Top