Approximately how wide must a roadway be in order for it to feasibly be reconfigured for center-running LRT tracks (and a single lane in each direction on either side)?
I tried to write a longer reply to the whole discussion, but always come down to what's basically a reiteration of this.It's worth remembering that in order for LRT to be worthwhile, it has to represent a true upgrade over buses in capacity, speed, and/or reliability - and not a downgrade. On primarily surface routes, buses will usually be faster than light rail; they accelerate and decelerate faster, and they tend to have faster doors as well. You can mitigate that with aggressive TSP/preempts at signals, but generally light rail will not be faster until you have a mostly grade-separated right of way. That's why the Mattapan Line still exists, and it's why a GL branch to Nubian could be workable - the downtown tunnel is enough of a speed and reliability boost to make trains preferable to buses.
So the question becomes, what do you gain by switching to LRT? You do gain a one-seat ride from Ashmont and Milton to points on BHA south of Talbot Avenue, and if you spend for the Warren Street tunnel you probably save a few minutes on trips using that segment. You get a one-seat ride into the subway from however far south your subway service goes - but going too far will sack your reliability.
I posted this map some time ago, but it's relevant again. This is what a fully built-out BRT network in the RDM area would look like. Relatively low cost ($15M/mile for a full rebuild of BHA), puts almost everywhere within a 5-10 minute walk of high quality transit, hits almost every stop on an enhanced Fairmount Line.
View attachment 23720
Yes, an LRT corridor through Blue Hill Ave and Seaver St to Jackson Square (and possibly to Hyde Park and Heath St) is very feasible... But it essentially duplicates the 29, which is already the least used route in the entire 2x3 matrix.
It's d be cool if you could get a one seat ride to Ruggles from close to ANYWHERE on the Commuter Rail... North South Rail Link cough cough.I very much agree with your overall point and with @The EGE as well.
That being said, I think this sentiment regarding the 29 is worth closer examination. Of the Nubian/Egleston routes, the 29 is the only one that doesn't through-run to Ruggles during regular service. For comparison, the 22 from Ashmont sees 970 passengers alight at Jackson Sq, but then another 340 alight at Roxbury Crossing and another 600 at Ruggles -- i.e. the extension to Ruggles doubles the number of passengers. We see similar pattens on the 23 and 28: the 23 drops 1,400 passengers at Nubian and then 1,500 at Ruggles, while the 28 drops 1,640 passengers at Nubian and 1,500 at Ruggles.
The 2,250 weekday riders figure on the 29 in the Better Bus Profile would probably be increased if there was actually a one-seat ride to Ruggles, which clearly sees high demand from across the southern half of the city.
The 29 also sees much lower frequencies:
View attachment 23743
Those frequencies are worse than they were pre-Covid, but even then, it was not great and not on the same level as the 22, 23, or 28:
View attachment 23744
Additionally, the T claims that the new bus lanes on Columbus (currently from Jackson Sq to Walnut Ave) saves 4 to 7 minutes of travel time. That's actually pretty significant. The 29 saw widely diverging travel times (both scheduled and actual), but all were in the range where even a 5-minute improvement would have been non-trivial:
View attachment 23745
The 29's route between Mattapan and Jackson is about 4.2 miles long. The current bus lanes are .75 miles long. I can't find any proposals for bus lanes on Seaver St, but bus lanes on Blue Hill would add another 2.5 miles of dedicated lanes to the 29's route -- 3.25 miles out of 4.2, or 77%. While I want to be cautious to avoid an apples-to-oranges comparison when estimating travel time improvements, on the face of it the Blue Hill lanes would be over three times as long, and if we do indeed see similar improvements of 4-7 min per .75 miles, that's a potential improvement of 12-21 minutes, which would be a huge reduction in travel time.
(Just to sanity check -- that reduction would put us somewhere in the range of 20 mins for the 4.2 miles between Mattapan and Jackson, which comes out to about 13 mph, which isn't totally unbelievable. The 28 going inbound takes ~40ish minutes to go a similar distance -- closer to 6 mph.)
Finally, it's possible that an extension -- whether BRT or LRT -- from Ruggles (or even possibly from Jackson) directly to Longwood would see greater demand than the current 29 does. Check out this OnTheMap analysis of employment locations for residences within the walkshed of the 22 and 29 from Egleston eastward:
View attachment 23748
View attachment 23750
From which we can see that Longwood is indeed a significant employment center, with Nubian and Boston Medical Center coming in a strong third place.
Now combine all those ideas together:
All of which is an extremely long-winded way to say: the current ridership of the 29 may not be a reliable indicator of potential demand for a BRT or LRT line via dedicated lanes on Columbus and Seaver. By providing speedier, more frequent and more reliable service directly to Ruggles and/or Longwood, there's reasonable evidence to suggest that some rider journeys might be diverted.
- increased frequencies
- significantly faster journeys
- one-seat to Ruggles, and/or
- one-seat to Longwood
Here's a properly crazy pitch:
Connect the Fitchburg Line to the inner Worcester by building ~2 miles of track alongside I95 in it's reservation between Brandeis and Auburndale. Then either send Red out from Alewife, or a fork of the Green from Porter to Brandeis, with the other fork doing Watertown. If it's green, the Fitchburg Line loses it's direct access to Porter/Red, but not the connection, either via Red or Green if a new transfer station /P&R is built at Brandeis. A new layover can live at the Aggregate Industries site or the Weston's shooters club - or, if you're even more ambitious, quad track the 95 leg to get a connection to Riverside. Getting to Brandeis, there's 4 crossings you'd have to treat for Red, including the 2 difficult ones downtown, and the 90/95 interchange for the CR.
Waltham communities would see improved frequency and better service to downtown, which if Red would create direct connections to Kendall and South, and GLX to Porter and beyond no longer needs to undercut the Porter CR platform or address the Fitchburg Line RoW - it can just take it over. Meanwhile, the West Cambridge/Alewife area split by the current CR tracks and MoW facilities between the Watertown fork and red service can be given over for redevelopment and street grid connection. (Though I would probably still want the T to keep a RoW through it all.) Green is probably more appropriate however - a series of LRT spaced infills at West Cambridge etc would greatly help with walksheds and eliminate crossing touches, and take them out of CR zone fares.
The Fitchburg Line would however also be competing for limited SS and Worcester Line slots, losing access to the BET and it's downtown layover in the process. and it's newly relocated tracks in Somerville would be somewhat wasted.
Why is Green connecting Kenmore and Harvard better than Blue (if Blue was extended to Kenmore)? Simply because there's not enough room in the BU Bridge/Throat area and GJ has to be Green?
If Blue is extended to Kenmore via the Riverbank subway, trips from Harvard on a further extended Blue Line would not be able to access the core of Back Bay. A Green Line extension would presumably have the option to run from Harvard via GJ or into the current Green Line tunnel, which would take pressure off the Red Line and Park Street transfers.Why is Green connecting Kenmore and Harvard better than Blue (if Blue was extended to Kenmore)? Simply because there's not enough room in the BU Bridge/Throat area and GJ has to be Green?
Another reason: IMO, Blue as an HRT line has greater potential west of Kenmore than to serve the Kenmore-Harvard link. In a Crazy Transit Pitches sense, plenty of areas would love the capacity and speed of an HRT line to downtown: Allston, Brighton, Watertown, Newton, etc. The exact direction and alignment are obviously up in the air, but most of them are more valuable than Harvard, which can be comfortably handled with a GL branch.Why is Green connecting Kenmore and Harvard better than Blue (if Blue was extended to Kenmore)? Simply because there's not enough room in the BU Bridge/Throat area and GJ has to be Green?
Providence & Worcester
- Worcester - Woonsocket - Providence
- Probably a strong commuter rail line, but not quite regional rail
- I could see arguments both for extremely limited stops (i.e. only Woonsocket and Pawtucket) and for lots of local stops (e.g. Millville, Uxbridge, Millbury) and lots of options in-between; given the size of the major cities served, an inaugural limited stop service could probably be viable
Only an hour? Not bloody likely. The NYNH&H/B&M State of Maine intercity train, the last passenger train to make this route, took 70 minutes to do Worcester-Lowell in 1941 with only 1 intermediate stop in Ayer. And took 2:05 to do Worcester-Haverhill. The curviness of the route is extreme and keeps the maximum speed limit with proper maintenance pretty low. There's definitely no way you'd do any better than the SoM '41 schedule adding local stops. And you'd have to temper your expectations for the audience if 70-90 minutes is the best-case. It would probably be a ton faster to run a coach bus on I-290/I-495 linking the cities than it would to try a rail solution for practical commutes. Geometry is an arch-enemy with what you have to work with here.Lowell & Worcester
- Worcester - Clinton - Ft Devens - Ayer - Lowell
- Commuter rail
- This journey would take about an hour -- extending to Haverhill looks good on paper, but would not be useful for Haverhill-Worcester journeys, which would almost certainly be faster via Boston
- Timed transfers at both ends going in both directions opens access to reverse commutes to New Hampshire, Springfield, and widens the reach of Boston commutes from places like West Boylston and Westford
The Southbridge Branch is a nonstarter given the meager size of the towns served and the roundabout routing of the ROW that avoids most village population centers en route. You'd honestly do better with buses on the local state highways if Webster, Dudley, Southbridge, etc. are in need of service. But WOR-NLN did survive until Amtrak A-Day in 1971, so the mainline corridor is a potentially underrated one. ConnDOT has frequent study flirtations with it, since P&W maintains the tracks to Class 3 (60 MPH passenger). The upgrade costs would be pretty modest, so it can thrive within the ridership of those towns.New London & Worcester
- Worcester - Webster - Plainfield - Norwich - New London (for maximum crayoning, add a branch to Southbridge)
- Commuter rail, though getting close to intercity rail when traveling end to end
- Connects a surprising number of destinations, including Groton (via ferry), Connecticut College, the casinos, and a string of towns running south of Worcester, and connects those communities to the Northeast Corridor via transfer at New London
First of all...we can't even get support for a Berkshire Flyer-type operation linking Boston to Greenfield via Athol because the riderships past Gardner are so microscopic. Substituting Worcester as hub is even more of a nonstarter. MA 2 simply isn't that congested until you hit Fitchburg, so running buses to Wachusett pretty much captures the market and makes it a matter of what cities you pair Wachusett with.Holden Branch
- Worcester - Holden - Gardner - Athol - Orange - Greenfield (with possible reverse branch running Greenfield - ... - Gardner - Fitchburg - Ayer - Lowell)
- Commuter rail, enabling 2-seat intercity journeys via transfer at Worcester
- There is a small but genuine Gardner-Worcester commuter market (and the reverse branch would be to serve the Gardner-Fitchburg/Leominster and Fitchburg-Lowell markets); moreover, traveling from the Northern Tier to Boston via a transfer at Worcester to a high-speed service will likely be faster than traveling via Fitchburg
- This one is a crazy transit pitch, but I think works better than expected
The abandoned ROW is non-landbanked, so that's impractical on its face. You could easily enough do Worcester-Fitchburg on the active CSX line by banging a left at the Ayer wye, though at 45 minutes Worcester-Ayer (more the more you add local stops on the branch) + 30 minutes to Wachusett the times get intolerably long. Worcester-Fitchburg/Gardner starts looking a lot like Worcester-Lowell in that the highways (in this case I-190/MA 2) are likely a ton faster than even the bestest rail routing money can buy. So if there's a commuter market here, it's one you prove by running high-quality coach bus service between city pairs like with Worcester-Lowell.Fitchburg & Worcester
- Worcester - Stirling - Leominster - Fitchburg
- Commuter rail
- The ROW is long abandoned and/or converted to rail trail through Stirling and through Leominster, but a quick scan suggests limited encroachment; there is a sizable Worcester-Fitchburg market, IIRC, so this could be worthwhile
- Note that the curve at Ashburnham puts a significant time penalty on journeys that travel west from Fitchburg; I'd still argue that sending Northern Tier service via Holden would end up being a better solution for Gardner and points west
Look at it this way:Knowledge Corridor
Hartford Line
- Springfield - Holyoke - Northampton - Greenfield
- Commuter rail, enabling 2-seat intercity journeys via transfer at Springfield
- Supplemented by East-West Rail services that continue north to Brattleboro, east to Worcester and Boston, and south to Connecticut
- I'd suggest that at least one peak trip should through-run to enable one-seat supercommutes between Northampton/Holyoke and Boston (e.g. arriving in Boston around 10 or 10:30); at other times, platform transfer enables intercity journeys
- For "maximum crayoning", some commuter rail trains running into Springfield could be extended to Ludlow or Palmer to turn those into commuter rail "satellites" -- Palmer might get an intercity stop, but Ludlow and East Springfield wouldn't, so it would be up to commuter rail service to pick up the slack; could be useful to enable Palmer-Hartford commutes, for example, but the east-west density is so much less than the north-south density
- Springfield - Hartford - New Haven
- Commuter rail or regional rail
- Basically today's service but enhanced, probably with some through-runs to Northampton and maybe Palmer
Hugely impractical. The B&A is hugely curved by the corkscrew of the Westfield River, so it would be a considerable time chew over relatively short distance to institute local service here. You can make do when Amtrak is hopping nonstop between Springfield and Pittsfield, but not with any quantity of local stops. As for reaching Downtown Westfield?...the Canal Line junctions pointing away from Springfield, so it would take a reverse move to get there at all. Plus the ROW south of the B&A is now a popular trail. Total nonstarter. Westfield's in the PVTA bus district. Pulse up those frequencies to Springfield Union using the straighter road network, and implement an express bus flavor if time needs to be saved on the trip.Westfield Branch
- Springfield - Westfield
- Commuter rail
- This one is hard and I'm not sold on it; Westfield State is far from the ROW and downtown Westfield is on the other side of a river
- Due to the track layout at Springfield, it would be hard to tie a commuter service to Westfield into a longer corridor
Reasonable pitch is to just do the real Cape Codder restoration: a Penn Station-Hyannis May-September weekender train. It was very popular until its schedule was unnecessarily fucked with.And while on the subject of Massachusetts non-commuter rail services:
Cape Codder v2
Newport Train Supersized
- Providence - Taunton - Middleboro - Cape Cod
- Long distance rail
- Cross-platform transfer to NE Regional services from New York
- Boston - Fall River - Newport
- Long distance rail
- Think CapeFlyer
- Points against: need for mode change or mode mixing since Aquidneck Island is unlikely to be electrified; needs to compete with NE Regional + Bus From Kingston and NE Regional + Ferry From Providence
- But it looks cool on the map!