The way I see it, the future of the Grand Junction falls into three eras.
First Era: Mainline
This is the current state. All trains must be FRA-compliant, all trains must always get priority at crossings, no compromise.
This is the state in which the Grand Junction will remain as long as the T needs to run the Readville Switcher for car moves to/from Boston Engine Terminal (the MBTA Commuter Rail Maintenance Facility, but tbh “BET” is more fun to say). Some combination of a new southside maintenance facility, improvements to the Worcester <> Ayer Line, and ultimately the North South Rail Link could lead to the Readville Switcher’s retirement.
Multiple studies have found that the Grand Junction is not ideal for frequent mainline rail. In particular, the frequency of gate closures, and the resulting impact on Cambridge’s major arterials, puts a cap on how often trains can pass through while still providing reasonable windows for traffic to cross over the tracks. In a lot of ways, the unsuitability of frequent mainline rail is unsurprising — there is a reason railroads stopped running at street-level in city downtowns (and Kendall is unmistakably a city downtown).
Mainline rail on the GJ is further constrained by where it can go beyond the boundaries of the corridor. Absent major capital projects, mainline trains can either run a) west to West Station, Boston Landing, and beyond, b) into North Station, and c) north toward Sullivan by crossing the Fitchburg tracks and running along the BET. The latest study outlines the operational impacts of all of these, and they are non-trivial. The lowest-impact option is (a), where there is sufficient space and extant track to run Grand Junction trains on separate tracks from the current Worcester trains.
The most valuable option would be a connection to Sullivan (requiring a new commuter rail platform to be built there, non-trivially), to create a reliable connection between Kendall and the bus hub at Sullivan, relieving pressure on the Orange Line as riders transfer at Sullivan rather than Downtown Crossing. Beyond needing a new platform at Sullivan, this option would also impact operations on all northside lines except Lowell, as the Grand Junction service either takes up “slots” or block tracks when crossing.
All of the above notwithstanding, the only solutions for retiring the Readville Switcher are large-scale/systemic — meaning that any study whose scope is limited to the Grand Junction itself is all-but-inevitably going to conclude that mainline rail will continue to be the only viable option.
Second Era: Semi-Permeable Light Rail
This is the state we have often pictured in ArchBoston’s Green Line Reconfiguration thread, and historically in various incarnations of the Urban Ring. (User F-Line to Dudley sketched out a version of this iteration here.) Here, the ROW has been transferred from mainline rail to light rail, but largely remains at grade, with most of the current grade crossings remaining. F-Line drew a short elevated overpass at Mass Ave to eliminate one of the busiest crossings.
There are a range of possible designs for this era, but they are united by a tolerance for interaction with the urban environment, and even some level of “cooperation” with it. Light rail vehicles are small enough, light enough, and able to stop quickly enough that the current grade crossings can be treated more like traffic intersections; the train would still get priority, but would be able to yield to cars and pedestrians when needed.
Beyond the corridor, light rail brings its own pros and cons. Both ends of the GJ corridor are relatively near existing light rail, although connections would be non-trivial to build. To the north, mainline ROW would need to be claimed for LRT (or greenfield ROW constructed) to reach Sullivan (to say nothing of needing grade-separated crossing of the Fitchburg Line etc). To the south, Comm Ave is relatively closeby, but, as I’ve written about before, going beyond Comm Ave or Kenmore requires a lot more work. However, the flexibility of light rail as a mode means that there are more options for less-invasive builds that could extend the corridor further without requiring 100% grade separation.
Third Era: Sealed Transit
This is the heavyweight version of Grand Junction transit. Tracks and stations completely removed from the street environment, with no interference from or deference to pedestrians or automobiles. Completely grade-separating the tracks allows you to run more trains, more frequently, faster, and, as noted, potentially with automation.
“Light metro” is sometimes suggested as an option for this (I presume in the style of the Docklands Light Railway) but it’s worth noting that this could also be traditional heavy rail, or various permutations of light rail. This decision will likely be governed by factors beyond the Grand Junction corridor, specifically the suitability of those different modes for the various outlying corridors. And it’s important to note that “Sealed Transit Along Grand Junction” does not necessarily entail “Sealed Transit” along those outlying corridors. Again, this is where light rail’s flexibility matters: for example, a Grand Junction el might be paired with light rail running Highland Branch-style along the Northern Strand Community Trail, which is largely but not totally grade separated. DLR-style trains and heavy rail would be limited to connecting to corridors that have themselves also been “sealed”.
I have frequently voiced skepticism about “sealed” builds for the Grand Junction corridor, but I do want to be clear that it could form part of an excellent transit service. Kendall has a very successful east-west transit line; a north-south counterpart — assuming it can be adequately integrated into the “Sullivan node” and the “Kenmore/BU/Longwood node” — would likely also be very successful.
Implications
First and foremost, I think any Grand Junction proposal needs to be understood in the context of these three eras and their structural requirements and prerequisites. To “graduate” out of the first era, the Readville Switcher problem needs a clear solution that is treated as part of the GJ proposal. To graduate into the third era, sufficient confidence about the appropriate mode for going beyond Sullivan and beyond Kenmore(ish) is required (which rapidly balloons the scope of any such proposal).
Second, the importance of future-proofing becomes clear. The “best fit” for the Grand Junction is going to evolve as the overall system evolves; light rail becomes viable once some tunnel or maintenance facility is built Somewhere Else; sealed transit becomes a stronger bet once (for example) a path for a Longwood subway comes into focus. The scenario that makes me the most anxious is some sort of mainline rail build that inadvertantly closes off possiblities for later conversion to LRT
Finally, and maybe most importantly, I think the various moving parts here necessitate phased improvements. If we have to wait for everything to be squared away before starting planning/construction, this project will never start. A strategy for Grand Junction transit needs to incorporate flexibility baked directly in to the proposal itself. If we can’t build a Sullivan <> Kendall <> Kenmore line in one fell swoop, then we need to be able to some partial version of it as a viable first step with multiple potential paths to the full-build from there.