Green Line Reconfiguration

[Aprehensive_Words]

In fairness, the T's predecessors did have such a conversion in mind when they built Kenmore, and the T gave it a study in the early 1970s, via extending the Blue Line down a narrow side street in Beacon Hill cutting a diagonal across the Common to meet up with the Central Subway near Arlington (broadly similar to the proposals contained in the 1926 report, but bypassing Boylston).

 

[Riverside]

I have more thoughts I might share later, but the other big piece (in my opinion) is the question of grade-separation. With light rail, you have some flexibility for things like grade crossings at streets and even sometimes street-running. With heavy rail, you have no wiggle room at all, especially if power comes from the third rail. A light rail project can be significantly less expensive if one or two costly grade separations can be avoided.

 

[kdmc]

the red line is already absurdly crowded during rush hours and even with better frequency more people working downtown is not a solution.

I would argue that RL crowding is less-so a function of downtown density and more-so a function of slow zones, a missing orbital line, and missing link between North and South stations. Unless they hop on a bus, anyone north of the river needs to travel downtown in order to reach Back Bay, LMA, Fenway, or the southern commuter rail network.

Plus, there are a lot of upsides to having a super-dense downtown. More street activity, less room for automobiles, short distances between offices and food/coffee/nightlife. In fact, with post-pandemic WFH, Boston is currently feeling the negative effects of *too much* vacancy downtown.

 

[Riverside]

additionally, i dream of a system with a few commuter trunk lines with many light rail/streetcar feeder routes, and the green line is essentially perfectly set up to showcase this for implementation in other communities.

You might find this piece I wrote about the B & C Lines interesting. It sounds like you are describing what I call the "Kenmore Model"; I am critical of that model in the linked piece, but I do think there is value to it.

I also imagined a system where more examples of the Kenmore Model were implemented, which I think is similar to what you're describing.

 

[Hooledeli]

sorry i was busy and forgot to check the forum

You might find this piece I wrote about the B & C Lines interesting. It sounds like you are describing what I call the "Kenmore Model"; I am critical of that model in the linked piece, but I do think there is value to it.

I also imagined a system where more examples of the Kenmore Model were implemented, which I think is similar to what you're describing.

View attachment 49962

this is awesome, yeah its kinda similar. heres my "plans" for a future dorchester neighborhood LR system in dark orange
(ignore the stations theyre all wrong)

i have something similar in mind for harvard square which i made a google maps for here:
MBTA LRV/BRT transit routes
unfinished but my idea for both these is to make trolleybus routes with dedicated ROWs to eventually become LRV if ridership is high enough
i know this stuff is very expensive but considering the mattapan trolley gets abt 2.5k daily ridership i think this is all perfectly reasonable (eveyr line i made follows only bus routes with at least 3k daily ridership by 2013 numbers) also i am very stubborn about trolleybusses over battery electric idc how many wires the mbta tears up

I have more thoughts I might share later, but the other big piece (in my opinion) is the question of grade-separation. With light rail, you have some flexibility for things like grade crossings at streets and even sometimes street-running. With heavy rail, you have no wiggle room at all, especially if power comes from the third rail. A light rail project can be significantly less expensive if one or two costly grade separations can be avoided.

i mean to be fair the commuter rail crosses roads at a lot of places at speed with crossing gates, and regardless the b branch could be sunken very easily because the avenue is so wide, the D branch can be rebuilt to use the inner track, and the c branch can keep trams and use the loop, again all theoretical i dont think heavy rail is very important

I would argue that RL crowding is less-so a function of downtown density and more-so a function of slow zones, a missing orbital line, and missing link between North and South stations. Unless they hop on a bus, anyone north of the river needs to travel downtown in order to reach Back Bay, LMA, Fenway, or the southern commuter rail network.

Plus, there are a lot of upsides to having a super-dense downtown. More street activity, less room for automobiles, short distances between offices and food/coffee/nightlife. In fact, with post-pandemic WFH, Boston is currently feeling the negative effects of *too much* vacancy downtown.

im all for density, and i know Boston has lots of space to grow. i just think we need to start now with polycentrism because otherwise well end up liek new york with ten billion parralell metros that all go to the same place because all 20 million people work in two places (no new york hate its just a bit claustrophobic)

In fairness, the T's predecessors did have such a conversion in mind when they built Kenmore, and the T gave it a study in the early 1970s, via extending the Blue Line down a narrow side street in Beacon Hill cutting a diagonal across the Common to meet up with the Central Subway near Arlington (broadly similar to the proposals contained in the 1926 report, but bypassing Boylston).

not gonna go into depth but another ide ai had ive seen literally nowhere is swinging it southeast from storrow drive to a station at copley and back bay, then following the road before going under the highway then replacing the B branch of the red line, giving the ashmont branch the higher ridership it deserves while still serving braintree and stuff

i dont know how to quote people from the last page, theratmeister said that boston is alreayd pretty polycentric, and i kinda agree honestly, doesnt change my mentality just another point against my GR to heavy rail idea. more orbital routes are absolutely needed

on the last page henryallen explaine dsome things about the capacity of light rail. i dont have a response but i do want our GR rolling stock to move towards something like the budapest suburban rail system
hungarian suburban rail
they have these big full low-floor cars that i think look really cool and i think we should use things like this instead of just making longer and longer streetcars. we can still use something like the current rolling stock for the C branch, cuz its the only one thats really like a streetcar

riverside, your post about the B and C branches was very informative, a difference of opinoin i have is that the B branch with its less road crossings and less stations is closer to the D branch, again with how wide the avenue is i think itd be easy to sink it a bit without a tunnel and eliminate basically all but a couple of the road crossings basically just making ir a more urban D branch (i like this idea asthetically so i may be biased sorry)

sorry if this post is too long i havent hit a word limit idk if theres an unspoken rule about post length or seomthing

 

[TheRatmeister]

also i am very stubborn about trolleybusses over battery electric idc how many wires the mbta tears up

You're in good company here.

i mean to be fair the commuter rail crosses roads at a lot of places at speed with crossing gates,

This works fine on small rural roads with infrequent service, not so much in urban areas with rapid transit frequencies. Crossing gates take ~20 seconds to come down, are down for an extra ~10 seconds on each side of the train passing, and then take another ~20 seconds to come back up. So now a train crossing takes 1 minute instead of ~10 seconds. If our goal is 5 minute headways that would mean 40% of the time the gates would be either down or going up/down.

and regardless the b branch could be sunken very easily because the avenue is so wide,

Yes, although there might need to be some extra considerations made due to the steep grade on parts of Comm Ave as well as the crossing of I-90. The real question is whether a subway for part/all of the route would provide enough improvement to the service over TSP/Intersection elimination/A bit more stop consolidation to justify the cost which would obviously be an order of magnitude or two higher. My suspicion is no, but I could be wrong.

the D branch can be rebuilt to use the inner track

This is harder than you might think. This would require diving the D branch under the C branch and then probably rebuilding the B branch incline to be further west so a junction could be added. I'm not even sure that this is possible due to the grades you'd need to navigate the changes in elevation.

and the c branch can keep trams and use the loop

This means cutting off a good chunk of Brookline from downtown and adding a transfer to all trips on the C. Yes, it's a cross-platform transfer so it's not that bad, but it would still be wildly unpopular, and not without reason.

i just think we need to start now with polycentrism because otherwise well end up like new york with ten billion parallell metros that all go to the same place because all 20 million people work in two places (no new york hate its just a bit claustrophobic)

Boston really isn't growing that fast at the moment, and as @kdmc said, Downtown is already chronically empty, and once again, we have plenty of job centers outside downtown. This isn't an urgent cause that needs an immediate course correction on a planning level.

not gonna go into depth but another idea i had i've seen literally nowhere is swinging it southeast from storrow drive to a station at copley and back bay, then following the road before going under the highway then replacing the B branch of the red line, giving the ashmont branch the higher ridership it deserves while still serving braintree and stuff

Somehow this is not actually that circuitous of a route, it works out about the same as the current RL. Again the concern is cost. The plan for the RL is 3 min headways on the trunk, so every 6 minutes to Ashmont/Braintree. There's also the fact that lots of Ashmont's ridership comes from bus transfers, so moving these out to the Fairmount Line would reduce the load on this branch and make those headways much more acceptable.

on the last page henryallen explaine dsome things about the capacity of light rail. i dont have a response but i do want our GR rolling stock to move towards something like the budapest suburban rail system
hungarian suburban rail
they have these big full low-floor cars that i think look really cool and i think we should use things like this instead of just making longer and longer streetcars. we can still use something like the current rolling stock for the C branch, cuz its the only one thats really like a streetcar

That's an interesting proposal that is basically half-high floor light rail. If this can be done without completely rebuilding all the stations with track crossings it becomes quite appealing, otherwise it's probably just easier to build a new line rather than try and get more capacity out of the GL.

 

[Teban54]

Welcome to archBoston, @Hooledeli!

I'll respond to the overall discussion on heavy rail conversion of GL later (I'm putting my thoughts in a blogpost, and I hope it won't disappoint). But I want to address one point quickly:

as @kdmc said, Downtown is already chronically empty, and once again, we have plenty of job centers outside downtown.

Strong disagree.

I analyzed job numbers in various employment centers a few months ago, and I redid the analysis below to only count private jobs: (as public jobs often end up being concentrated at a single location, like city hall, regardless of where the jobs actually are)

• DTX/Financial District/Chinatown (roughly bounded by Tremont St, Stuart/Kneeland St, the waters and State St): 116,423
• Back Bay (between Arlington St and Mass Ave, south to SW Corridor/Columbus): 64,653
• Seaport: 40,810
• Kendall/MIT (roughly bounded by Charles St, Windsor St, Mass Ave and the river): 59,950
  • Some offices on the western edge of this region may be closer to Central than Kendall
• Longwood Medical Area (cutoff slightly south of Francis St): 59,952
• Harvard (roughly bounded by Ash St, the river, Flagg St, Putman Ave, Ware St, Francis Ave, and Mellen St): 25,064
  • The figure is almost entirely from Harvard University itself, which also includes the Allston campus
• BU-Kenmore (from Harvard Ave to Mass Ave, north of Comm Ave-Pike): 21,353
  • The figure is almost entirely from BU itself

I've tried a few other places (Fenway, NEU, BUMC, UMass, Logan Airport...), but no single tangible area gets sufficiently above 10K.

I personally think of North Station/MGH, Back Bay and Seaport as three legs of "greater downtown". Taken together, the area below hosts 308,535 jobs. This is not to forget that the center of it (Financial District) accounts for 38% of the polygon's jobs, despite only accounting for 14% of the area (though, to be fair, the polygon includes a lot of water). In contrast, the 4 aforementioned job centers outside of downtown only host 54% of jobs compared to greater downtown.

For funsies, the following polygon -- most of the "denser" metro area except "greater downtown" -- has 539,159 jobs. This is 174% of greater downtown's jobs, but over 3000% of its area.

Boston does not have a "chronically empty" downtown by any means. The need for circumferential transit does NOT imply a lack of demand for radial transit. If anything, it's the reverse: additional transfer opportunities are needed to provide relief for downtown transfer stations, which already see huge ridership from both entries and transfers. While it's not to the extent of "we need to start now with polycentrism" as @Hooledeli said, the distribution of jobs above does not particularly strike me as "pretty polycentric" either: it's mostly just 2 local maxima.

Even arguments like WFH do not undermine demand for downtown Boston. My incomplete analysis of 2023 vs. 2019 ridership showed that downtown stations' Covid recovery rates are not substantially different from other job centers like Kendall and Harvard. In Jan-Oct 2023, South Station and Downtown Crossing have the #1 and #2 ridership systemwide in terms of station entries, which do not account for transfers between rapid transit lines.

It's not like transit service downtown has no room for improvement, either. Seaport is still criminally underserved. Surface transit (such as buses) sucks. South Station is still only served by the overburdened Red Line despite its extreme importance. Even Back Bay could arguably use more routes to areas not served by GL and OL, which is possibly why the 1 bus is so popular.

 

[HenryAlan]

For funsies, the following polygon -- most of the "denser" metro area except "greater downtown" -- has 539,159 jobs. This is 174% of greater downtown's jobs, but over 3000% of its area.

This is really cool. What tool are you using to produce these maps, and what data sets can it use besides employment?

 

[Teban54]

This is really cool. What tool are you using to produce these maps, and what data sets can it use besides employment?

This data was produced using US Census OnTheMap, a rather specialized but easy-to-use interactive tool with employment data. Its most unique feature, one that I didn't use in the analysis above, is that it shows origin-destination employment data.

For example, here's the distribution of workers who hold private jobs in Longwood Medical Area:

And here are the job locations for residents in Cambridge near Central Square working in private sectors:

For each of these queries, the map also gives a demographic profile of the queried workers, such as age, race and income.

In addition to this and querying the total number of jobs in an area (which was used in my previous analysis), one last functionality is to easily query the number of employed persons in any polygon, which can approximate population (especially when comparing multiple areas). On that front, OnTheMap's role is not quite unique, as many other population datasets exist and are more accurate. But I think it's by far the most customizable - you can enter just any area, instead of being restricted to municipal boundaries or having to manually add up a bunch of census tracts.

OnTheMap doesn't have data outside of employment, but there's a large array of data from the US Census that are available on multiple official and third-party interactive tools. The most exhaustive is probably the official US Census data viewer, which lets you view basically every dataset, but it's quite cumbersome to use. I showed several examples of these in an earlier comment in the context of Jamaica Plain, and I'll replicate one of them here. This map shows the proportion of residents that commute by car:

More advanced users who have knowledge of GIS software, such as ArcGIS and QGIS, can download the geometric datasets and perform more sophisticated analyses and visualizations (tutorial). I happened to be learning QGIS the past weekend, and here's a spoiler for a map I'm making for the blogpost:

Spoiler: Work-in-progress map spoiler

This map shows population density overlayed with the rapid transit network. I'm only showing a small section of it, leaving the full map for the blogpost. Can you guess what the colors for rapid transit stations represent?

 

[Tallguy]

Does anyone know what the minimum row for the GL. I guess technically, from outer track to outer track

 

[TheRatmeister]

Does anyone know what the minimum row for the GL. I guess technically, from outer track to outer track

I believe around 24 ft not including barriers between tracks/road lanes.

 

[Hooledeli]

Made this green line map based off of the ideas you've all discussed. Wanted ur thoughts before I make a higher effort digital one (this map does not include the radial lines)

The only idea I did that I haven't seen discussed is branching the Waltham green line off just north of magoun square and following the red line for a short bit (aboveground) there's room for both the train and the community path and trees on either side

An extension past union square could also go left under Washington and Kirkland Street to Harvard

I have replaced the green line with 10 separate lines here, but it could easily all be branches of the green line, I haven't read much of the discussions around renaming stuff so idk

 

[TheRatmeister]

Here's a quick table I put together that I'll just call: "How much time do the GL surface branches lose to long dwell times?"
GLX and subway stations generally have a dwell time of around 45 seconds, but because the surface stops don't have all-door boarding their dwell times here generally longer. Here's how much time that costs for each branch by direction, and what the average speed of each branch would be with all dwells capped at 45 seconds. (For some surface stops like Dean Rd it would make more sense to cap at 30 seconds, but let's just apply a blanket 45 second cap for now since it doesn't require a complicated methodology.) All data comes from Transitmatters, collected between 4/3/24 and 5/4/24.

Branch/Direction	Total time lost from extra long dwells	Speed without long dwells
B Outbound	~2:15	~9 MPH (Up from ~8.3 MPH)
B Inbound	~2:30	~8.8 MPH (Up from ~8 MPH)
C Outbound	~1:00	~9.7 MPH (Up from ~9.1 MPH)
C Inbound	~1:30	~9 MPH (Up from ~8.3 MPH)
E Outbound (Just Brigham Circle-Northeastern)	~1:30	~9.2 MPH (Up from ~7.5 MPH)
E Inbound (Just Brigham Circle-Northeastern)	~1:00	~8.4 MPH (Up from ~7.5 MPH)

 

[Hooledeli]

Here's a quick table I put together that I'll just call: "How much time do the GL surface branches lose to long dwell times?"
GLX and subway stations generally have a dwell time of around 45 seconds, but because the surface stops don't have all-door boarding their dwell times here generally longer. Here's how much time that costs for each branch by direction, and what the average speed of each branch would be with all dwells capped at 45 seconds. (For some surface stops like Dean Rd it would make more sense to cap at 30 seconds, but let's just apply a blanket 45 second cap for now since it doesn't require a complicated methodology.) All data comes from Transitmatters, collected between 4/3/24 and 5/4/24.

Branch/Direction	Total time lost from extra long dwells	Speed without long dwells
B Outbound	~2:15	~9 MPH (Up from ~8.3 MPH)
B Inbound	~2:30	~8.8 MPH (Up from ~8 MPH)
C Outbound	~1:00	~9.7 MPH (Up from ~9.1 MPH)
C Inbound	~1:30	~9 MPH (Up from ~8.3 MPH)
E Outbound (Just Brigham Circle-Northeastern)	~1:30	~9.2 MPH (Up from ~7.5 MPH)
E Inbound (Just Brigham Circle-Northeastern)	~1:00	~8.4 MPH (Up from ~7.5 MPH)

What is the solution to this? Putting fare collectors at every door (I think they already did this they're just not on yet) has to trust people will use them, and if every station is rebuilt with fare gates there's nothing stopping people from walking on the tracks into the station

 

[TheRatmeister]

What is the solution to this? Putting fare collectors at every door (I think they already did this they're just not on yet)

The difference between the surface stops and the subway stops is all door boarding, so pretty much.

 

[The EGE]

All-door boarding at surface stops with proof of payment (POP) is standard on basically every light rail system except the Green Line. Yes, fare evasion is possible, but it's not actually the massive issue that Pioneer et al make it out to be. Systems that have converted haven't seen a massive increase in fare evasion - it turns out that most riders are basically honest. Trying to get the remaining few to pay (who end up, generally, being the least able to) probably costs your system more in increased operating costs than it saves in additional fares. It's no coincidence that the fare evasion story is pushed by Pioneer and their ilk who care more about getting more cops in stations than they do about transit efficiency or serving riders.

The benefits of all-door boarding are well established. San Francisco saw a 38% per-passenger reduction in dwell time on buses, which increased reliability by evening out dwell times.

https://www.sfmta.com/sites/default/files/agendaitems/2014/12-2-14%20Item%2014%20All%20Door%20Boarding%20Report.pdf