Teban54

Just another transit enthusiast

[Idea & Analysis] The Kendall-Back Bay Subway: From an Orange Line Re-route to the “Pink Line”

—

by

Teban54
in ,

TL;DR

  • People talk about reverse branching the Red Line via a new Mass Ave subway all the time. But if you shift the alignment further east, you now get another crazy idea: Reverse branching… the Orange Line?
  • Turns out, such a direct link between Sullivan, Lechmere, Kendall and Copley/Back Bay can be valuable in itself. I present two proposals that use it as a standalone route, the “Pink Line”.
  • Personally, I favor this alignment over Mass Ave: it hits the high-demand destinations at more central locations than Mass Ave does. Comparing the two alignments raises a question: How often are our imaginations limited by what exists today?
  • However, neither can replace a more conventional Urban Ring — which means such a subway proposal will likely remain too aggressive to be prioritized.
  • Nevertheless, this allows for a closer look at the roles of Back Bay vs. Kendall in the network, or even in the region as a whole. They have much more similarities than you might expect. Why is Back Bay often considered “extended downtown” nowadays, but not Kendall?

Background: The “Mass Ave Red Line” Proposal

You had probably seen this proposal before: Reverse branching the Red Line (RL) via Mass Ave.

The Red Line reverse branching proposal, which has decent popularity among transit enthusiasts.

The motivation is very natural and explains its popularity: Mass Ave, a major north-south road corridor in a region where north-south rapid transit connectivity is scarce, also happens to be on or close to the Red Line on both ends. So why not create an alternative route for Red Line trains to run under?

This is an example of reverse branching:

  • Typically, a rapid transit line runs in a single trunk in the “downtown” area where demand is at its highest and may sometimes diverge into multiple branches in the “outer neighborhoods” where demand is lower.
  • But the proposal above creates two reverse branches within “downtown”, one via Park St and one via Mass Ave. Each branch’s frequency is half of that of the common northside terminus, Alewife.

To be clear: Reverse branching is often a bad idea. RailsRoadsRiverside and Alon Levy have written about this in depth. Thus, I am NOT in favor of the Mass Ave Red Line proposal myself, both due to reverse branching and for many other reasons (detailed later).

Reverse Branching for Funsies: The Orange Line

But if we want to have some fun and play around with reverse branching anyway… Why stop at the Red Line?

Entering an even crazier proposal: Reverse branching the Orange Line (OL), via… Lechmere and Kendall!

My Orange Line reverse branching idea. [NOT A SERIOUS PROPOSAL]

Disclaimer: This is NOT a serious proposal. While some aspects of it are worth considering, I am NOT proposing to reverse branch the Orange Line exactly as such.

This idea is probably much less trivial than the Mass Ave one. But here’s the unserious, almost childish motivation:

The Orange Line makes a big C-shaped detour in downtown Boston between Sullivan Square and Back Bay, and it’s even more roundabout than the Red Line. What if we cut it straight?

The Sullivan – Lechmere – Kendall – Back Bay Subway

Once we look between Sullivan Sq and Back Bay, we see a few very, very interesting nodes… Lechmere and Kendall.

Geographical Alignment

Despite the lack of an underlying road network (unlike Mass Ave), these nodes are much better aligned north-to-south than you might expect:

  • Lechmere is almost due south of Sullivan. If you want, there’s even a rail right-of-way (ROW), the Grand Junction, that continues due south past Sullivan for longer than OL itself does.
  • Third St, the main north-south road corridor through Kendall Sq, ends just two blocks away from Lechmere’s station entrance (and one block from the bus loop).
  • Dartmouth St is one of the most prominent north-south roads through Back Bay, passing through Copley Square and both Green and Orange Lines’ stations. If you extrapolate Dartmouth St further north across Charles River… You end up in Kendall, almost dead on at the Red Line station and Third St!

So, if you want to be God and ignore construction costs… Why not build a tunnel straight through all this?

In fact, most of my map above follows existing roads, and where it doesn’t, it was mostly crossing either the Charles or rail yards. The only exception is going from First St to Third St in East Cambridge.

Strong Local Demands

Not only are these nodes nicely aligned, but they all have great prominence themselves:

  • Sullivan Square is the first major transfer hub north of downtown Boston. Today, 11 bus routes bring residents from Somerville, Charlestown, Everett, Chelsea and more places to connect to the Orange Line.
  • Lechmere is not only the diverging point of both branches of Green Line Extension (GLX), but the area has also seen plenty of recent development. Cambridge Crossing is now home to many high-rise apartments and corporate offices.
  • Kendall Square is arguably the biggest employment center outside of downtown Boston, with around 70,000 jobs between tech, MIT and more. It hosts more jobs than Central Sq and Harvard Sq combined. The Red Line sees significant “reverse commute” demand from downtown Boston into Cambridge during AM peak, with a large amount attributed to Kendall.
  • Back Bay (including Copley Square) is what I would consider Boston’s “extended downtown”, if not part of the downtown itself. Besides its own 70,000 or so jobs (similar to Kendall), it’s also a vibrant center for shopping, dining, conferences and more.

Circumferential Connections

In addition to bringing people to destinations along the line, such a hypothetical subway can also play a crucial role in the network: connecting (almost) all other lines, and doing so outside the downtown core.

Boston’s rapid transit network is almost entirely hub-and-spoke in nature, with lines meeting in the downtown core (Shawmut Peninsula) and radiating outwards. This is often not ideal for people traveling between neighborhoods, or (arguably more often) to employment centers away from the core.

Central Square (Cambridge) is 1.7 miles from Copley Square as the crow flies, but the Red and Green lines carry you through 3.3 miles — almost twice as long!

This is why most of the world’s largest rapid transit systems typically have either a ring route or overlapping services forming a ring. Even in Boston, the idea of an “Urban Ring” has been touted around forever, and even went through official studies in the 2000s.

While this Kendall-Back Bay proposal doesn’t look quite like a conventional Urban Ring, it achieves several similar objectives. More comparisons will be made in later sections.

Towards a Standalone line: The Pink Line

Even without seriously advocating for reverse branching either RL or OL — which I don’t — we can nevertheless still use the core alignment of the Sullivan – Lechmere – Kendall – Back Bay subway to get slightly more sensible ideas.

The Red-Orange Marriage: A Crazier “Red X”

I’ll first start with a proposal that’s still in the “insane” category, but less so than reverse branching.

Let’s think of the more standard (non-reverse) branching, where a line splits outside of downtown towards different neighborhoods. Two such cases are relevant here:

Red Line branches, and the “Red X” proposals

The Red Line has two branches to the south. This has always limited the frequencies that each of Ashmont and Braintree can receive, especially in recent years when slow zones and shortages plagued the branches with 20+ min headways. Even though the Red Line Transformation project will eventually improve branch headways to 6 minutes, it will always be twice as long as that of the main Red Line trunk: every 3 minutes.

There have been various fantasy proposals to “save” the Red Line branches, by diverting one of them into a new right-of-way (ROW) north of JFK/UMass. This idea is often called the “Red X”: Turning the line from a λ shape into two independent X-shaped lines, such that each of them can run every 3 minutes.

Most mainstream Red X proposals usually have the new trunk head to downtown Boston. This includes the textbook Red X by F-Line to Dudley, which essentially builds a Congress St Subway from South Station to North Station.

Orange Line branching proposals

While the Orange Line doesn’t have any branches today, there have nevertheless been countless ideas for a new OL branch among transit enthusiasts and even real-world advocacy groups.

The most popular idea is an Orange Line branch to Everett. North of Sullivan Square, the branch would diverge onto the Eastern Route (the Newburyport/Rockport Line’s ROW) to Sweetser Circle, from which it either heads north under Everett’s Broadway or continues east to Chelsea.

I myself have never been fond of branching the Orange Line as a long-term measure, and the primary reason is, again, frequency.

  • The ongoing Orange Line Transformation project will upgrade both its signals and rolling stock to support a frequency of “only” 4.5 minutes. Creating branches without any other signaling upgrades would result in 9-minute frequencies for Malden, which I feel is unacceptable given its status as a dense suburb and bus transfer hub.
  • The Lower Mystic Regional Working Group report included an Orange Line branch as Alternative 8. It also advocated for improving OL’s frequencies to every 3 minutes, which requires “an additional 78 cars […], a new signal system, upgrades in power supply, and an expanded maintenance facility.” Even with such investments, a Malden branch would still only see trains every 6 minutes, a downgrade from 4.5 minutes.
  • Not to mention, all of this assume rush-hour headways. Frequencies during off-peak hours and especially on weekends would likely become intolerable on each branch, as anyone from Dorchester and Quincy can tell you.

However, a new rapid transit line that does not merge with the Orange Line seems perfectly reasonable, if not necessary. So…

If there’s an existing Red Line branch to the south, and a frequently touted Orange Line “branch” to the north… Why not join the two branches with a new trunk through downtown? So far, this is still within the realm of reality for standard Red X proposals (if not the most common belief of what the Red X should do).

But here’s where the “crazy “insane” part comes in: Instead of going through the downtown core, why not connect them via the Kendall-Back Bay subway?

My idea of the Pink Line from Everett to Braintree, via the Kendall-Back Bay subway, Roxbury, and taking over the RL Braintree Branch. [NOT A SERIOUS PROPOSAL]

Disclaimer: This is NOT a serious proposal. While some aspects of it are worth considering, I am NOT proposing to reroute the Braintree Branch exactly as such.

The reason why this is not serious should be obvious: It’s too much of a detour for Quincy and Braintree, particularly for those seeking downtown.

  • Most riders probably want to head to the downtown core. Even though Copley can be seen as the “extended downtown”, I doubt it has the same draw.
  • Riders seeking the original Red Line will have a forced transfer at Fields Corner. I haven’t found any way to make the transfer seamless: it will probably be a tedious elevated-to-underground transfer no matter what. In contrast, more traditional Red X proposals often allow cross-platform transfers at JFK/UMass with track modifications.

(Somewhat surprisingly, distance and “roundaboutness” aren’t actually the main issue. Today’s Red Line spans 7.86 miles between Kendall and North Quincy, whereas my Pink Line takes… 7.95 miles. Longer, but not by much.)

Having said that, there are many valuable and/or fun aspects to consider out of this (otherwise insane) proposal:

Why radial line to Everett?

Continuing the Kendall-BBY subway to South End

Rapid transit to Nubian, not just a streetcar

The rest of Roxbury: Where neither Fairmount nor Nubian do justice

The “L-hook”: From Grove Hall to Neponset (and Quincy)

Down to Earth, or Down to City Point

One final proposal, that’s also the most realistic (though still expensive):

A more realistic Pink Line proposal from Everett to City Point, via the Kendall-Back Bay subway and South End.

For the first time in this post, I’m actually comfortable with suggesting this as a concrete, “halfway-realistic” proposal. This is not to say that it will happen: the “revised” Pink Line will still be expensive (see below). In fact, it’s still well above the “budget” that I’m usually comfortable with.

But back to the proposal itself. The main change from the “crazy Pink Line” is freeing it from responsibilities that are not the most well-suited for the Kendall-Back Bay subway: namely, both the “Red X” (absorbing one of its branches) and the Nubian-Warren line.

Instead, this now allows the Pink Line to head to another neighborhood that’s hungry for its own rapid transit: South Boston (Southie).

South Boston’s density and current transit access

As a recurring theme in this post… South Boston is yet another neighborhood that qualifies for better transit, in theory. Its density is quite comparable to Cambridge, only outshined by the absolute densest neighborhoods in the entire region.

While Southie is technically served by two Red Line stations, they’re at the literal corners of the neighborhood, with up to 3/4 of their walksheds blocked by highways and rail yards.

Most Southie residents need to transfer to buses, which have surprisingly strong demand: Routes 7 and 9 are among the most frequent and most heavily used bus routes that are not designated Key Bus Routes, which is even more impressive given their bad midday and weekend headways. Both are expected to become Frequent Bus Routes under the Bus Network Redesign.

Unique circumstances: Is grade-separated transit really the best?

However, this may be a case where “does place X need heavy rail” goes beyond just the demographical statistics alone.

I argue that a few unique factors are at play, which may change the calculus:

  • South Boston’s employment patterns are very downtown-centric, especially Financial District-centric. It may be one of the most downtown-centric neighborhoods in metro Boston. A direct, shorter ride to downtown may be preferable over a transfer, especially an “out of the way” one.
  • South Boston is much closer to downtown Boston than most other neighborhoods. Thus, time savings from achieving higher speed between the neighborhood and downtown become less impactful.
  • Since it’s somewhat geographically isolated, South Boston is primarily a destination, not a transfer hub (unlike Harvard, Nubian, Sullivan etc). There’s less of a need to carry “suburban” riders from further away into downtown quickly.
  • West Broadway seems busier than Dorchester St with more commercial activities. Meanwhile, anything from Seaport (Summer St or L St) misses out on a huge chunk of South Boston’s population.

These factors may have the following implications:

  • A route through W Broadway – E Broadway, transferring to the Red Line at Broadway station, may be the most preferred alignment. It has the best coverage (unlike Seaport-South Station) while also being in the “right” direction towards downtown Boston (unlike Andrew).
  • Compared to a fast service that makes fewer stops and requires more walking, it’s plausible that that residents may prefer a “local” service with more stops closer to home.
  • Anything that’s more expensive than bus improvements should probably offer a one-seat ride to (some parts of) downtown.

This already implies that the Pink Line may not actually be the most ideal solution for South Boston. It doesn’t go to the more “conventional” downtown, only to Back Bay. The number of stops it can make will likely be limited, if built as heavy rail.

That said, my intention of routing the Pink Line as such was merely to explore the options, not to claim decisively that it’s the best.

Comparing Kendall-Back Bay to Other Alignments

For the rest of this article, forget about the exact northern and southern alignments of the Kendall-Back Bay subway, beyond Sullivan and South End. Let’s instead compare it to other parallel rail corridors.

Kendall-BBY vs. Mass Ave

Here’s my opinion: The Sullivan-Kendall-Back Bay hits places that are more centrally located, and with higher demand, than the Mass Ave alignment.

Job Distributions

Transit Ridership

Anecdote on the 1 bus: Are riders using it as a compromise?

Personally, these factors sum up my gripes with a Mass Ave service: It barely hits or barely misses several major destinations, and feels like a half-baked way to serve them. Kendall, Back Bay, Kenmore-Fenway, Ruggles, LMA… You name it. A route that’s either just to the east, or just to the west, can serve them more effectively.

  • The only major nodes that Mass Ave directly hits are MIT, Boston Medical Center and South Bay Center (Newmarket). But MIT’s campus is already very elongated and doesn’t favor either Mass Ave or Kendall strongly, while the other two are way too far south and fundamentally a different discussion.

This makes me wonder: How much of the current successes and future fantasies of Mass Ave are simply because the road exists? Are our imaginations limited by what’s already there?

Obviously, if you’re talking about BRT or surface LRT with grade crossings, then Mass Ave is the best we can get. But the moment you consider grade separation and building a tunnel under the widest part of Charles River, following Mass Ave may not even net you that much cost savings.

Any river crossing here — minus reusing the existing BU Bridge — will be expensive anyway. So if you’re paying that cost, why not do it right, or at least think about if that’s doing it right?

Kendall-BBY vs. Urban Ring

I’ll keep this briefer than the previous comparison:

I do NOT think of a Kendall-Back Bay subway as a replacement of Urban Ring. An actual Urban Ring service, in my view, should serve destinations further west: Kenmore/Fenway/BU/Allston, Longwood Medical Area, etc. In a God-mode world, these two services should co-exist.

What an Urban Ring should do — and why Kenmore-BBY doesn’t replace it

Rethinking Back Bay’s role — and even Kendall’s

The discussion above touches on some interesting aspects of Back Bay, though.

In a vacuum, Back Bay alone is more prominent than Kendall alone or LMA alone. So why has Back Bay not been involved in most Urban Ring discussions so far?

Distance from downtown may be a reason: Copley Square is around 1 mile from Downtown Crossing (DTX) in straight-line distance, while Kendall is 1.4 miles away. But that alone doesn’t explain Seaport — another node that’s often involved in Urban Ring — which is also 1 mile from DTX.

Personally, the “real” underlying reason for me is:

Back Bay feels more like an “extended downtown”, rather than “outside of downtown”.

The Role of Geography (Lack of Water Crossings)

Back Bay vs. Kendall: A Historical Perspective

This comparison gives us another lens to reconsider not just Back Bay, but also Kendall.

Traditionally, I had always thought of Kendall as an analog of LMA. Both are major employment centers (~70k jobs) that are clearly “outside of downtown”.

But now, it seems that Kendall and Back Bay have a lot in common:

  • Their distances to downtown are roughly on the same scale.
  • Historically, they were both intended as “stepping stones” to outer hubs. Even though both stations almost didn’t exist, the transit access they eventually got has undoubtedly played a huge part in their developments.
  • Today, they host a similar number of jobs, while taking on varying degrees of non-employment activities (commercial vs. educational).

If I had thought of Back Bay and even Seaport as “extended downtown”… Then why is Kendall typically considered “outside of downtown”, not “extended downtown”? Is it simply because of the elephant in the room — Charles River?

(This geographical factor definitely affects our perceptions today. But I think it might have also affected real-world decision making, specifically in planning of the High Spine.)

Even though this would seem to argue for adding Back Bay to Urban Ring considerations, I maintain that it might not be a good idea for other reasons. Many other factors are at play: strength of nodes further out, cost, etc.

Regardless, these parallels between Kendall and Back Bay are natural explanations for a fantasy proposal linking them together. The idea didn’t come out of nowhere.

Why All of This is Fantasy

As I’ve mentioned repeatedly: This entire proposal (even the revised Pink Line) is much less realistic than my typical ones.

I am usually a somewhat conservative crayonist. Instead of assuming everything can be brute forced with a tunnel boring machine (TBM) at unlimited cost, I like to keep cost, engineering and budgets in consideration. A perfect service plan that costs tens of billions of dollars is worse than a less desirable route that actually gets built. However, this post is one where I intentionally let go of those constraints.

The situation gets even more dire for the Kendall-Back Bay subway, in particular:

Cost

Long sections of subways will always be expensive. But some are more expensive than others.

Tunneling under and around Charles River

Any underwater crossing in this part of Charles River needs to be at last 2,200 ft long, which is already complex

But the challenges may actually extend to areas under land. F-Line to Dudley once stated that, due to soft, low-quality landfill in most parts of Cambridgeport south of the Red Line’s ROW, any subway there has to be designed almost like a cross-harbor tunnel to mitigate lateral movements. It’s not clear how much this applies to the Boston side of the river, but we do know that almost the entire Back Bay neighborhood was landfilled.

This affects all of:

  • The Kendall-Back Bay subway
  • The Mass Ave subway
  • A subway under Grand Junction (much more popular idea)

F-Line thinks that a Mass Ave subway can quite plausibly “cost just as much as NSRL. And certainly more than any other individual transit project ever.”

I’m not knowledgeable enough to verify these claims… But they don’t paint a bright picture. Keep in mind, the NSRL tunnel is roughly as long as one between Kendall and Back Bay stations, without going under water. The official MassDOT NSRL Feasibility Assessment in 2018 estimated its cost between $12 billion and $22 billion.

(If anything, burying a tunnel between Kendall and Dartmouth St may be easier than dead under Mass Ave. You’ll have a lot more space on open water for immersing tubes.)

Dartmouth St in Back Bay

Tunneling under these streets is less of a concern. But even if you ignore possible landfill issues under Dartmouth St, the Back Bay side can still give you serious political concerns, especially if you attempt the cheaper cut and cover (C&C) method that requires digging it up entirely.

(One advantage of C&C here, to be fair, is that Dartmouth St has lots of cross streets. This might make closing short stretches of the road less disruptive.)

Then there’s the challenge of putting a station in Copley Square, surrounded by tons of historical buildings: Old South Church and Boston Public Library immediately to the west, and Trinity Church not far to the east. Given earlier woes in 2018 when the Green Line station’s accessibility work damaged Old South Church, further construction will likely face at least political pushback, if not engineering.

Third St in Cambridge

The Cambridge side of the route may have fewer rich people, but Third St is also narrower (50 ft building-to-building in some sections).

There’s also the turn from Third St to First St (or at least Second St), where cutting through a block with a TBM is unavoidable, especially for HRT (which doesn’t have the luxury of allowing tight turns like the Green Line does).

If anything, I’m wondering if the longer alignment of McGrath – Land Blvd – Binney St – Third St would be better for cutting down costs.

To be clear: A Mass Ave subway faces most of the same challenges. Its only significant advantage is avoiding construction in Copley Square and crossing under blocks, but those may not be enough to keep its cost down.

Priority

This is probably the bigger concern: Neither Kendall-Back Bay nor Mass Ave are among the most urgent — nor most valuable — alignments here.

If a new Charles River crossing west of Longfellow Bridge were to be used for rapid transit, it would almost certainly be for Urban Ring. Despite Back Bay’s importance in itself, I have to think that any real-world Urban Ring proposal will be further west, with a greater focus on Longwood Medical Area — and rightfully so.

Even among alignments for new downtown trunks, neither rank particularly highly. If exactly one (non-NSRL) north-south downtown trunk is built, which is aggressive to begin with, the Congress St subway seems like a much better choice — especially for its most likely recipients to the north, Chelsea and Everett.

Lastly, these cross-Charles tunnels may not even be the most valuable “cross-harbor tunnels”. Remember that a Charles crossing near MIT needs to be at least 2,200 ft long? For “only” 700 ft more, you can cross from Seaport to Logan Airport. Such a tunnel would add much more unique, irrepleaceable value to the network than a cross-Charles tunnel.

Closing Thoughts

This idea started as a completely insane exercise: the Orange Line analog for the popular proposal of reverse branching the Red Line. Yet, it actually uncovers some interesting ideas and thought-provoking questions, at least in abstract route planning.

This is not the first time I’ve thought about the Kendall-Back Bay subway, but it’s the first time I’ve analyzed and reasoned about it in great depth. It made me rethink the roles of both Kendall and Back Bay in the transit network, and more broadly, in the region. The idea also allowed for a more rigorous comparison with the Mass Ave subway, leaving a question that is worth considering for everyone:

How often are our imaginations limited by what exists today? Not just the existing street grids (as is the case for Mass Ave), but also rail ROWs, and even other aspects such as developments and demand patterns?

(As a side note, the question of routing the southern half of the Pink Line also touches on a broader topic: How many new rapid transit routes south of Boston may be desirable? How should we feed them into trunks through downtown Boston? These questions may be addressed in a future post.)

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *