I'm finally reading this post after almost a full week. Excellent work!
Thanks,
@Teban54! Great feedback, thoughts below:
1. The redesign extends the 90 further east from Assembly to Wellington and Chelsea, likely to replace the Wellington-Chelsea link currently achieved by 112 and providing one-seat ride from Chelsea to Assembly Row. With your suggestion moving the 90 to Union Sq, how would this link be maintained? Perhaps a new route that runs Chelsea-Wellington-Assembly-Sullivan, at 30 mins or better?
Candidly, this was an oversight on my part, an artifact of the particular method I used for this project. The good news is that the method I used to calculate the savings from cutting back the 90 from Wellington to Sullivan does
not count the mileage from Wellington to Chelsea. So if we wanted to just leave in place a Chelsea-Wellington 30-min-or-better route, that'd be no problem.
Now, as for Chelsea-Assembly, Chelsea-Sullivan, and Chelsea-Somerville...
(Need to hold myself back from diving too far down the rabbit hole here.)
In my opinion, the simplest solution is to reroute from Chelsea-Wellington to Chelsea-Sullivan via Alford Street. That requires filling a deficit of 0.6 route-miles (at thirty-min freqs). I'm optimistic that mild frequency adjustments across the system could fill this deficit, but if not, it could be
filled via shortening the T96. I'd also note that this restores a connection between Chelsea and Sullivan, which is lost in the Redesign and is definitely
not rectified by the Redesign's 90.
I'm wavering in my conviction here, but I still ultimately feel that the best way to serve Assembly is via high-freq shuttles from Sullivan. Shuttles would offer door-to-door service to stores, and can connect to a vastly greater number of destinations via a transfer at Sullivan. I'd argue that this is superior to a one-seat-ride 30-min-or-better along a single corridor.
(Case in point, transfer via a high-freq shuttle at Sullivan would make a journey from Mystic Ave to Assembly much more feasible than the Redesign's strategy; under the Redesign, you need to coordinate a transfer between two mid-freq schedules, while a shuttle would allow you to focus on the one mid-freq schedule and be confident that there will be a shuttle waiting for you.)
But... (he says, teetering on the edge of the rabbit hole.)
The Redesign's 90 is -- I guess? -- an attempt to provide Chelsea a link to the Orange Line and to the northeast quadrant. This seemed odd to me; they have high-freq corridors radiating out from Everett to Glendale, Ferry St, Wellington, Sullivan, Chelsea, and Revere -- surely one of those would link Chelsea with the Orange Line? The answer, of course, is no: the T104 runs from Malden to Ferry St to Chelsea to Wood Island. This makes some sense, because Malden - Everett - Chelsea/Wood Island more or less forms a straight line, as do the other high-freq corridors: Wellington - Everett - Revere on the T110, Sullivan - Everett - Glendale on the T109.
This trio of services centered on Everett consists of one radial service and two criss-crossing circumferential services; that's why the Chelsea high-freq corridor ends up being routed
away from its nearest Orange Line connections, and why (I suspect) the Redesign's 90 was included.
I modified their map to highlight the radial vs circumferential routes:
(Climbing out of the rabbit hole...)
Blah blah blah, the point I'm faffing around here is that I think there are a few things worth reconsidering in this segment of the network. I'm skeptical about converting the Ferry St corridor into a circumferential, and likewise Beach St in Revere. I'm skeptical about the lack of connection between Chelsea and Sullivan, though I do note that (I believe) every transfer opportunity that would be available at Sullivan is still served elsewhere by the Redesign's 90. A more systematic rework might be able to address the general issue of what to do with this half of the Redesigned 90.
Coming back to the most straightforward solution: keeping the Chelsea-Wellington segment in place. This still leaves out connectivity to Assembly... but honestly if we can swing a shuttle system from Sullivan, it wouldn't be
that much more difficult to run it to Wellington as well. And, as mentioned, there's at least one way to generate the 0.6 route-miles needed to shift the destination from Wellington to Sullivan, which I maintain would be a viable (if not better) solution.
2. If you're cutting the 87 to Sullivan-Mystic Ave, how would you serve Turkey Hill, which is currently served by 67 (to be withdrawn) and 87 in the redesign? Extending your 90 from Arlington Center to Turkey Hill might be feasible, but creates a deficit again.
Though I did not make it explicit (and did indeed omit it from my map), my intention was to have the Turkey Hill segment reassigned from the 87 to the 90, which would ultimately run Turkey Hill - Arlington Center - Clarendon Hill - Davis - Highland Ave - Union Sq. Because the Turkey Hill segment would need to be reassigned anyway, I did not include it in the 3.3 route-miles surplus gained from shortening the 87.
3. One additional benefit of amending T101 to Ball Sq is that it almost recreates the Powder House Sq-Sullivan high-frequency corridor, especially for the Orange Line riders from Powder House Sq (OSR via 87 in the redesign, eliminated in your proposal).
Yes -- in an earlier version of this post, I also advocated rerouting the T101 in this manner;
I relocated it into the appendix as an "additional option" in order to keep focus on the major changes. But I definitely think it would be a good idea. As mentioned above, shortening the T96 to Davis frees up about 1.3 route-miles -- half of that could go to creating the Chelsea-Sullivan route I described above, and the other half could support this reroute.
4. Your proposed 87 has nothing in common with the current 87. In fact, your proposed 90 is basically the current 87 but cut back to Union Sq. For aesthetic purposes, I would call the (Turkey Hill?)-Arlington-Union route 87 and give the Mystic Ave route another number, maybe 82?
The numbering is a mess at this point, haha. The "current" 87 vs the "redesigned 87/proposed 87" vs Riverside's "suggested 87/revised 87"... it's real mess! For the purposes of this discussion, I've just tried to keep the designations used by the Redesign as much as possible; in some cases, because I've basically swapped entire routes, it creates the disconnect you're describing here. I figure the whole numbering system will be reviewed and tweaked prior to deployment, but I didn't want to create yet more confusion here by getting into that now.
5. I still wonder what's the best use of T39. The current 39 is a beast in both frequency (6-7 mins at peak) and capacity, which is definitely overkill for Somerville Ave (assuming T39 keeps the current 39's frequency and fleet), but might be overkill even for Highland Ave and Clarendon Hill. But to avoid adding too much complexity, the best option might be to simply short-turn some trips at Union Sq during peak hours.
Yeah, this is a real problem, and one I intentionally maintained as "out-of-scope" for this endeavor. (I actually drafted a section on the T39, but excised it for the final draft -- could be a post on its own.) There are a
number of issues and open questions around the T39; maintaining the idea of a "full frequency" T39 all the way to Clarendon Hill seemed like a "maximum possible build", so I figured I'd just go with that to be conservative in my estimates, and simplify the whole analysis.
But the point you raise gets at something I briefly mention in my appendix:
the Redesign avoids using short-turns at all, so blatantly and consistently that it clearly was an intentional design choice. But there are legitimate uses for them, and I think you've outlined one possible example. As I
noted in my appendix, there are a number of stretches where a few extra buses made available by short-turns on the T39 would be useful.
But I think there's a larger conversation that needs to be had about the T39; it's probably the boldest proposal in the Redesign, and I think is emblematic of many of the exciting elements of the Redesign. But I also think it presents downsides and challenges, and really needs to be examined wholistically.
Thanks again for the feedback! I will try to update the post soon to clarify the points you've raised here.