Transit via the Grand Junction Corridor | Cambridge and Boston

[nortor]

There have been a few media write-ups about Cambridge's recent Grand Junction transit feasibility report, but Cambridge Day's (from yesterday, 10/15) seems among the most comprehensive:

CRA sees a Grand Junction passenger service connecting North and future Allston stations - Cambridge Day

A historic rail line running through one of its densest business corridors could offer an attractive opportunity to run a new stream of trains through Cambridge, connecting two of the MBTA’s most ambitious long-term projects.

www.cambridgeday.com

I love that the thumbnail image is a BEMU render.

These are the relevant quotes about looking like Green Line trains. They don't seem very well considered.
Page 3-7:

Developing rail transit connecting commercial and residential hubs throughout Greater Boston will require introducing an emerging service concept in the region: Urban Rail. While this service would be operated with trains that look like Green Line trains, the equipment would be FRA compliant, thereby allowing it to operate on existing Commuter Rail tracks ... While this concept would be new to the region, there are numerous examples of Urban Rail operations throughout the United States and internationally. Urban Rail trainsets are typically shorter in length than traditional commuter rail consists, making them more appropriate for an urban setting. They often feature frequent service (i.e., headways of 15 to 20 minutes) and equipment that operates on either combustion of liquid fuels (e.g., diesel) or electricity.

Page 5-29:

It is assumed that the Grand Junction vehicles would be the same that the MBTA acquires for its proposedurban rail service as part of the Rail Vision project. Figure 67 shows renderings of a possible EMU or BEMU vehicles based on the Stadler KISS EMUs.

So, just regional rail trains. Not sure where "look like Green Line trains" comes in.

Also, from the report, page 4-15:

Within eastern Cambridge, the Grand Junction service would serve as a distributor and circulator transitservice for local trips along the alignment, operating at regular frequencies throughout the day. If 15 or17.5-minute headways are assumed on the West Station-North Station service, alongside 30-minuteservice between West Station and North Shore communities, the combined headway for new service within the City of Cambridge would be 10 or 12 minutes, respectively. Densely developed eastern Cambridge is home to major employment, retail, universities, and services, similar to a typical metropolitan area’s central business district / downtown. An analogue to what service along the Grand Junction would mean or “feel like” to those in the City of Cambridge would be urban streetcars. Such operations provide frequent fixed-guideway service that circulates and distributes riders among downtown activity centers .As shown in Table 37, daily ridership estimates for the eight American streetcar systems with lengths less than three miles mostly fall in the range of 2,000 to 4,000 daily riders. Average use across these systems yields approximately 2,100 daily weekday riders. This imperfect daily average of streetcar ridership was assumed for Grand Junction travel within eastern Cambridge. This simplistic analogous estimate does not take into account elements such as employment or residential density, fare, service frequency, nor does itc onsider the nature of the ridership. Based on the CTPS model data, intra-eastern Cambridge trips would grow by 12 percent between 2022 and 2040, equating to a future year total of approximately 2,350 weekday riders within this market.

This is so dumb that I'm not sure how seriously to take the rest of the report. As they themselves note, "this simplistic analogous estimate does not take into account elements such as employment or residential density, fare, service frequency, nor does it consider the nature of the ridership". Well yeah, obviously! So then why are they claiming a future year total of 2350 weekday riders out of thin air? The streetcars they are averaging (Atlanta, Tampa, Memphis, KC, Dallas, Salt Lake City, Milwaukee and DC) vary in typology, but they're completely different from this (even when only considering local transit)! Many are 2010s downtown circulator routes which have been criticized for poor service, slow speeds, and limited destination connections. An urban rail service in Cambridge serving Cambridgeport, MIT, Kendall, and East Cambridge stations as they propose might have higher or lower ridership, but it would be fundamentally different in character since it would be only 4 stops with presumably higher average speed than mixed traffic streetcars. This level of analysis shouldn't be acceptable in an official report.

They add later in their caveats section:

The analogous assumption for intra-Cambridge ridership, based on an average daily total from eight representative American streetcar systems, is simplistic, based only on system length. Other key area and system characteristics were not considered, such as: density of development along the corridor; right-of-way type (fully or only partially exclusive); running pattern (single direction loop versus bi-directional); fare; headways; and primary trip purpose (commute, shopping, tourism). Not examining these salient system elements has resulted in a usable but imperfect estimate for the intra-Cambridge market.

No, that's just as good as making a number up.

 

[ritchiew]

This is so dumb that I'm not sure how seriously to take the rest of the report. As they themselves note, "this simplistic analogous estimate does not take into account elements such as employment or residential density, fare, service frequency, nor does it consider the nature of the ridership". Well yeah, obviously! So then why are they claiming a future year total of 2350 weekday riders out of thin air? The streetcars they are averaging (Atlanta, Tampa, Memphis, KC, Dallas, Salt Lake City, Milwaukee and DC) vary in typology, but they're completely different from this (even when only considering local transit)! Many are 2010s downtown circulator routes which have been criticized for poor service, slow speeds, and limited destination connections. An urban rail service in Cambridge serving Cambridgeport, MIT, Kendall, and East Cambridge stations as they propose might have higher or lower ridership, but it would be fundamentally different in character since it would be only 4 stops with presumably higher average speed than mixed traffic streetcars. This level of analysis shouldn't be acceptable in an official report.

Yeah, this was surprisingly sloppy for a report like this.

For the transit planners out there, how should they have gone about predicting ridership? Really practically, is there some kind of standardized software? Or a lot of formulas that have to be cobbled together for each unique project?

 

[HenryAlan]

No, that's just as good as making a number up.

The other flaw, is that none of those examples are for lines that are tied in to a much larger and sophisticated network. Yes, DC Metro is great, but it doesn't really interact with the streetcar. Maybe the other services would be useful for intra-Cambridge riders, but that completely ignores the numbers feeding in to the line from the terminus transfer options.

 

[bigpicture7]

Yeah, this was surprisingly sloppy for a report like this.

For the transit planners out there, how should they have gone about predicting ridership? Really practically, is there some kind of standardized software? Or a lot of formulas that have to be cobbled together for each unique project?

The other flaw, is that none of those examples are for lines that are tied in to a much larger and sophisticated network. Yes, DC Metro is great, but it doesn't really interact with the streetcar. Maybe the other services would be useful for intra-Cambridge riders, but that completely ignores the numbers feeding in to the line from the terminus transfer options.

Huh?
That estimate notor cites isn't even an important part of the report. It's just from a subsection on "within eastern Cambridge" expectations. The report dedicates many pages to overall ridership. I am not saying the report is accurate or not, but this is a cherry picked critique from a 131 page document.

 

[F-Line to Dudley]

Page 5-29:

So, just regional rail trains. Not sure where "look like Green Line trains" comes in.

Not "just regional rail trains"...pigeonholed to literally one overcustomized variant of the Stadler KISS bi-level and nothing else. Because (page 5-31) they stupidly break from T standards and spec 8-inch platforms for all the GJ stops in a vain attempt to save money. The only vehicle on the planet that can accessibly board at East Coast-spec 48-inch platforms (North Station, Chelsea, Lynn, planned for West Station) and 8-inch low platforms is the over-expensive Caltrain-customization KISS bi-level EMU & BEMU. Nothing else in existence can physically do both, because there's no market on the planet for doing both. No commuter agency excepting perennially stupid/incoherent Caltrain was ever reckless enough to mix accessible boarding standards on the same system. So the CRA basically shot themselves in the foot on vehicle shopping options with that counterintuitive preference for low platforms (I guess you could say that's the only thing remotely Green Line-like here) when a full-high 250-footer doesn't appreciably bloat the station costs. And at least allows for the possibility of single-levels on a service that's ideal for single-levels.

You don't always get the vehicle you want in a procurement for a variety of reasons. If the T can't force-feed Caltrain-custom Stadler KISS bi-levels as low-bid winner, the project as-specced becomes physically impossible and they have to go back to the drawing board re-factoring all of the station construction costs for redesigned full-highs and their associated ramps. I can't even fathom what kind of cognitive dissonance thought that was a good decision.

 

[nortor]

I think the report is essentially just trying to show with a sort of order of magnitude calculation that the ridership of this section would be comparable to Commuter Rail Lines.

It should be recognized that this modeling effort uses estimates that almost wholly focus on commute trips, as opposed to discretionary activities (e.g., all trips not related to work duties). Ideally, one would use a travel demand model, calibrated to the eastern Cambridge region, for all purposes, with the latest demographic and transportation system assumptions. Nevertheless, the projections demonstrate that there is ample ridership potential based on the amount of transit commutes centered around eastern Cambridge from other parts of the Boston area that could benefit from a Grand Junction service .Initial base year ridership estimates for the aforementioned core Grand Junction service operated in conjunction with an additional service between Lynn and West Station that would only attract existing transit commuters, are anticipated to be greater than 2018 daily ridership on two commuter rail lines(Fairmount, Kingston/Plymouth) on the lower bound while upper bound ridership would be higher than recorded on all but four commuter rail lines in 2018 (Franklin, Providence/Stoughton, Worcester/Framingham, Newburyport/Rockport). When projected out to a 2040 horizon year, the lower bound daily ridership would be greater than recorded on five lines in 2018, while the upper bound daily ridership projection would still be greater than daily ridership recorded on all but three commuter rail lines.

It sounds like that travel demand model would be at least a better approach.

The other demand estimate components (for the 2040 estimate) are commute trips from: North Shore transferring at North Station (2666-5425), F/W line commuters transferring at West Station: (1075-2253), and Allston-Brighton commuters (472-1116). These are all based on current transit ridership on these lines, and the ranges are from half of people switching to a faster service to 100 percent of people (and then projecting to 2040 based on the Boston MPO's estimate of trip growth for each commuter market. These probably underestimate ridership if regional rail is implemented.

The sloppy estimate is actually a big portion of the projected ridership (2350 out of about 6563-11144).

Thee conclusions section for the core route in their demand analysis section gives some insight into the goals:

The lower bound condition would produce approximately 5,800daily riders in the base year, which is slightly less than the 2018 daily ridership for the Kingston/PlymouthLine (see page APP-6 of the Appendix for a table of ridership of existing Commuter Rail Lines). The upper bound scenario would produce 9,600 to 9,800 daily riders in 2022 for 15- and 17.5-minute headways respectively, which are slightly higher than the 2018 daily observations along the Fitchburg Line. This demonstrates proven ridership potential - were a Grand Junction service to open today (using the 2022base scenario as a proxy). Even if the large intra-eastern Cambridge market (2,100 base, 2,350 future) was removed from the analysis, the base year range would be 3,700 (lower bound) and 7,500 to 7,800 riders (upper bound) while the future year range would be 4,200 (lower bound) and 8,600 to 8,800 (upper bound). For both the future and base years, the analysis indicates that the lower bound daily ridership would still be considerably greater than the 2018 value for the least busy commuter rail line while the upper bound projections would see greater ridership than half the existing lines.... Even if the large intra-eastern Cambridge market (2,100 base, 2,350 future) was removed from the analysis, the base year range would be 3,700 (lower bound) and 7,500 to 7,800 riders (upper bound) while the future year range would be 4,200 (lower bound) and 8,600 to 8,800 (upper bound). For both the future and base years, the analysis indicates that the lower bound daily ridership would still be considerably greater than the 2018 value for the least busy commuter rail line while the upper bound projections would see greater ridership than half the existing lines.

So it's just trying to say that this has ridership similar to existing Commuter Rail Lines, even if you don't consider the "streetcar" ridership. This is fine, but I think it really undersells this corridor's utility for circumferential and non-commute trips, especially in a world with more frequent regional rail service.

So fine, yeah, the sloppiness doesn't affect that conclusion. But that's a pretty broad, and not especially useful, conclusion.

That estimate notor cites isn't even an important part of the report. It's just from a subsection on "within eastern Cambridge" expectations. The report dedicates many pages to overall ridership. I am not saying the report is accurate or not, but this is a cherry picked critique from a 131 page document.

The ridership estimates get reported uncritically by news outlets like Cambridge Day (not that they should have to nitpick the data) and become part of the public discussion around the project. The streetcar thing is just completely not based in reality, and is a large portion of those numbers. I think that's a problem.

 

[nortor]

Not "just regional rail trains"...pigeonholed to literally one overcustomized variant of the Stadler KISS bi-level and nothing else. Because (page 5-31) they stupidly break from T standards and spec 8-inch platforms for all the GJ stops in a vain attempt to save money. The only vehicle on the planet that can accessibly board at East Coast-spec 48-inch platforms (North Station, Chelsea, Lynn, planned for West Station) and 8-inch low platforms is the over-expensive Caltrain-customization KISS bi-level EMU & BEMU. Nothing else in existence can physically do both, because there's no market on the planet for doing both. No commuter agency excepting perennially stupid/incoherent Caltrain was ever reckless enough to mix accessible boarding standards on the same system. So the CRA basically shot themselves in the foot on vehicle shopping options with that counterintuitive preference for low platforms (I guess you could say that's the only thing remotely Green Line-like here) when a full-high 250-footer doesn't appreciably bloat the station costs. And at least allows for the possibility of single-levels on a service that's ideal for single-levels.

They say that the trains would be the same as the MBTA Regional Rail trains. But as you point out (and as the Green Line comparison demonstrates) they're obviously not thinking very hard about those things.

 

[bigpicture7]

I think the report is essentially just trying to show with a sort of order of magnitude calculation that the ridership of this section would be comparable to Commuter Rail Lines.

It sounds like that travel demand model would be at least a better approach.

The other demand estimate components (for the 2040 estimate) are commute trips from: North Shore transferring at North Station (2666-5425), F/W line commuters transferring at West Station: (1075-2253), and Allston-Brighton commuters (472-1116). These are all based on current transit ridership on these lines, and the ranges are from half of people switching to a faster service to 100 percent of people (and then projecting to 2040 based on the Boston MPO's estimate of trip growth for each commuter market. These probably underestimate ridership if regional rail is implemented.

The sloppy estimate is actually a big portion of the projected ridership (2350 out of about 6563-11144).

Thee conclusions section for the core route in their demand analysis section gives some insight into the goals:

So it's just trying to say that this has ridership similar to existing Commuter Rail Lines, even if you don't consider the "streetcar" ridership. This is fine, but I think it really undersells this corridor's utility for circumferential and non-commute trips, especially in a world with more frequent regional rail service.

So fine, yeah, the sloppiness doesn't affect that conclusion. But that's a pretty broad, and not especially useful, conclusion.


The ridership estimates get reported uncritically by news outlets like Cambridge Day (not that they should have to nitpick the data) and become part of the public discussion around the project. The streetcar thing is just completely not based in reality, and is a large portion of those numbers. I think that's a problem.

Fair points. My point was simply that there were many components to the study, not just that particular hand-wave. The study was done by a reputable firm. But the scope of what they're asked to examine and the sorts of press releases/exec summaries that get put out are another story.

 

[nortor]

Fair points. My point was simply that there were many components to the study, not just that particular hand-wave. The study was done by a reputable firm. But the scope of what they're asked to examine and the sorts of press releases/exec summaries that get put out are another story.

Agreed. The broader problem is that there should be a permanent regional body tasked with doing comprehensive travel studies and integrated transit project planning, design, and construction. Sadly that might be a pipe dream. These kinds of single project studies will, as you note, definitionally have an incomplete scope. The way this study precludes discussion of light rail without doing any sort of comprehensive cost-benefit analysis is a good example of the ways this approach is limiting.

In addition to the need to accommodate existing and potential rail uses by others (e.g., passenger rail equipment moves by MBTA and Amtrak, enduring or “legacy” rights of access for freight carriers), there is also the need to allow for the future implementation of the City of Cambridge’s planned Multi-Use Path Project, which is currently in design. In addition, only Federal Railroad Administration (FRA)-compliant vehicles are considered (i.e., the use of non-compliant vehicles is inherently out of scope). This intentional limitation is intended to prevent the advancement and further analysis of alternatives that would inevitably interrupt existing passenger and/orfreight services, and ultimately prove challenging to implement. Since legacy freight rights will endure until ceded by the carrier (CSX) and physical separation (explained in the next section) does not appear to be achievable within the existing ROW, the assumed need to avoid the use of non-compliant vehicles for a new passenger rail service effectively narrows the potential universe of alternatives even further (e.g., Light Rail Transit becomes less competitive since it would only be able to achieve FRA-compliant operations via temporal separation (without requiring additional ROW)).

If developed to meet the current fleet, implementation of LRT along the Grand unction would rely on the construction of OCS along the entire extent of the corridor. Putting aside the incremental capital costs associated with implementing OCS to enable LRT service, which would be considerable, such equipment would still need to be housed along a width-constrained rail corridor abutted by existing institutional, commercial, and residential structures. At a practical level, the ability for a Grand Junction-based LRT service to physically tie-in with existing transit services is relatively limited at either end. Headed north from Medford / Gore Street, the corridor passes through a sharp eastbound horizontal curve and beneath McGrath Highway / Squires Bridge en route to connecting to the Outbound track of the Fitchburg Line (an Inbound connection does not presently exist). In this area, the Green Line Extension descends from an elevated profile to an at-grade alignment in order to pass beneath the same roadway bridge en route to Union Square (Figure 27). Within the vicinity of where these tracks would logically connect, the Fitchburg Line trackage lies adjacent to, but substantially beneath, the Green Line’s elevated transition, with one intervening stub track (at-grade with Fitchburg Line). If the horizontal and vertical profile of the Green Line’s underpass segment (and the presence of the McGrath Highway Overpass that ultimately constrains them) are to be taken as fixed design elements, then a direct tie-in with the Green Line Extension should not be considered feasible. As the horizontal curve begins just north of Medford Street, providing a more indirect tie-in with the Green Line by turning west towards Union Square and then reversing direction, would require the acquisition of substantial land within the City of Somerville in order to house the associated railroad infrastructure.

 

[F-Line to Dudley]

They say that the trains would be the same as the MBTA Regional Rail trains. But as you point out (and as the Green Line comparison demonstrates) they're obviously not thinking very hard about those things.

On the contrary...they seemed to think very very hard about the most Rube Goldberg-esque possible solution to keep the station costs artificially low. Which is kind of scary given the effort put in to make that force-fit. It was just haphazardly compartmentalized to the point that they didn't realize they were tanking the procurement bid process and winding up with a vastly more expensive and self-limiting vehicle solution for their troubles.

Again...the mental gymnastics boggle the mind.