Regarding LRT on/under Warren St: I think this is a very complicated discussion -- both from a transit design perspective, and from a social-political one. Remember, in recent memory there was significant community opposition to the
"28X" BRT proposal. Am I saying that we should never propose BRT or LRT on Blue Hill Ave again? Absolutely not. What I'm saying is that an LRT line will run into many of the same objections, and a good proposal should have some way to address them.
Two transit design notes: first, and probably most importantly for a tunnel under Warren: where are you going to put the portals? (I talk about this the least, but it's also one of my biggest concerns.)
Second, a Mattapan-Park St LRT line would parallel a Fairmount Indigo Line at well-less than a mile away for its entire stretch south of Columbia Road; the infrastructure for a one-seat ride to downtown already exists -- does it need to be duplicated on a mode that will be slower?
Finally, it's worth recognizing that the transit network in Dorchester is unique in Boston. Three northern terminals (Nubian/Ruggles, Egleston/Jackson/Ruggles, and Forest Hills) are each linked to both southern terminals (Mattapan and Ashmont), with almost every route featuring very high frequencies all day. These routes overlap to form a lattice, in a structure that basically does not exist anywhere else on the T. (
I wrote at some length about the unique features of this network back in late 2020.) It is not trunk-and-branch.
Building a rapid transit spine down the center of this network would not have the same network-realigning effect that (for example) GLX to Medford is likely to have. Green-on-Blue-Hill handles one pair (Nubian-Mattapan, though only sorta), but you'd still need to keep your other routes pretty much exactly as they are: Nubian-Ashmont, Jackson-Ashmont, Jackson-Mattpan, Forest Hills-Ashmont, Forest Hills-Mattapan. And honestly, you probably need to keep your Nubian-Mattapan buses too, if you're going to do rapid transit spacing. Plus, there's the fact that a huge fraction of these riders actually go to
Ruggles, not Nubian -- my skim of the data suggests the split is roughly 50/50.
(And, by the by,
LRT from Nubian is not likely to be faster than staying on the bus and transfering at Ruggles.)
So... an LRT line from Park ends up serving:
- Blue Hill Ave & Warren St riders heading downtown who are able to walk to a rapid transit-spaced stop but are not able to walk to a more distant but also more centrally located Fairmount Line stop
- Blue Hill Ave & Warren St riders heading directly to Nubian
- Blue Hill Ave & Warren St riders transferring to a different bus at Nubian
But doesn't help
- any riders to Jackson Square
- any riders to Ruggles
- any riders who need to transfer to a different bus at Ruggles
- any riders to Longwood
- any riders to Forest Hills
This is my point about the Dorchester network being a lattice. It is true that the 28 corridor comes in first place in ridership... but all the other corridors are not far behind at all. The 28 corridor looks like it is a trunk line, but it really is more complicated than that.
I highly recommend reading the
Better Bus Profiles for the different routes we're talking about here. Another piece of the puzzle that comes through in those profiles: there is major ridership turnover on these routes -- many people aren't going to Nubian at all but are traveling locally from one point on the corridor to another. An LRT line will help some of those folks with a slightly faster ride, but many will continue to prefer the bus as its local stop spacing will get them closer to their destination (and likely will come more frequently than a Green Line train would anyway)
I don't mean to come across as hostile, so I apologize for that. Laying track in a newly-created median on Blue Hill Ave between Mattapan and Franklin Park Zoo? Sure, yes, from an engineering perspective that is probably easier than the work at Pleasant (easier from an engineering perspective but maybe not from a political one). Likewise, LRT tracks on Columbus and Seaver -- building on existing bus lanes -- would probably be similar.
A new cut-and-cover tunnel down a narrow street in a dense residential neighborhood, along a stretch of street that sees more buses than (I believe) anywhere else in the system that would all need to be diverted and delayed during construction, and many -- if not all -- of which would still be needed following completion of the LRT line... that is
much harder. Add that on top of the number of transit journeys that this won't really improve at all... my point is that, even if the tunnel is feasible (but again, portal locations), it's far from obvious to me that this is the best way to improve transit in Dorchester.
(Sorry for wall of text and for possibly hostile tone -- I don't mean to come across aggressively. Like I said, I just think there are a lot of layers here.)
This was my concern with the portals. I think it may be possible to squeeze it down closer to 400 feet, but it still creates a pretty large footprint.