MBTA Buses & Infrastructure

Also, on the 71/73, Load Factor is an important consideration in service delivery.
The ETBs had fewer seats than their diesel counterparts, which pushed the peak load factor higher.

1702496403285.png
 
Cambridge especially Harvard Square used to have a much more robust independent feeling—an endless number of small, independent businesses and shops and services that don’t really exist anymore. Now, there’s Harvard, of course, but other than that it feels very hollowed out, employment opportunity wise, unless you’re looking to bus tables or something similar.
Your mention of employment opportunities at Harvard reminds me of an analysis I've done recently, which looks at ridership at Harvard, Central and Kendall/MIT in Fall 2019:
Harvard Central Kendall subway riderships Fall19.png


The bottom line seems to be: Harvard and Central each seems to draw half as many employers as Kendall alone (i.e., people commuting to these stations during AM peak). Depending on your interpretation, this either speaks highly of Harvard and Central, or speaks highly of Kendall.
  • Harvard and Central are both residential stations. Kendall a lot less so.
    • AM Peak inbound ons: 2402, 2595, 739 (H, C, K respectively)
    • PM Peak outbound offs: 2951, 3613, 1275
  • All three are employment centers (somewhat surprisingly for H and C). But Kendall is the most major, having as many commuting riders as the other two combined.
    • AM Peak outbound offs: 2289, 1859, 4449
    • PM Peak inbound ons: 3396, 2296, 4916
  • Kendall also has commuters from RL North (Alewife-Central), which is again the sum of the other two stations. But Harvard and Central do draw some commutes from RL North, too, surprisingly.
    • AM Peak inbound offs: 1107, 1571, 2956
    • PM Peak outbound ons: 1380, 1747, 3613
Note that a large chunk of the residential ridership at Harvard (i.e., people commuting from Harvard during AM peak) comes from bus transfers, as illustrated with the 2015-17 ridership data below:

StationHarvardCentralKendall
Walk & cycle92551114516701
Drive, carpool, dropoff264287449
Transfer from bus1122751451008

The table alone indicates that 54% of riders at Harvard come from buses, but this ratio should be significantly higher if you also make the assumption that people working at Harvard walk to the station rather than taking a bus (i.e., they work near Harvard Square itself instead of an employment center along a bus route).
 
Last edited:
Note that a large chunk of the residential ridership at Harvard (i.e., people commuting from Harvard during AM peak) comes from bus transfers, as illustrated with the 2015-17 ridership data below:

StationHarvardCentralKendall
Walk & cycle92551114516701
Drive, carpool, dropoff264287449
Transfer from bus1122751451008

The table alone indicates that 54% of riders at Harvard come from buses, but this ratio should be significantly higher if you also make the assumption that people working at Harvard walk to the station rather than taking a bus (i.e., they work near Harvard Square itself instead of an employment center along a bus route).
Cool data. Im wondering about this last table, and whether the data are parseable by bus line? I would be very curious to see the breakdown by bus line.
 
Cool data. Im wondering about this last table, and whether the data are parseable by bus line? I would be very curious to see the breakdown by bus line.
The data in this table alone can't be separated easily, as the survey simply asked anyone boarding at Harvard what's their previous mode of transportation. However, you can break down bus ridership by route using another dataset on bus ridership per route and per stop. I did an analysis here that showed weekly ridership at Nubian, Harvard and Sullivan for all bus routes in Fall 2022, and here's the table for Harvard: (includes the Harvard busway, the 1's stop at Mass Ave @ Holyoke St, the 66/68/69's stop at Johnston Gate and Dawes Island, and Eliot St @ Bennett St)

Harvard bus riderships Fall22.png
 
Also, on the 71/73, Load Factor is an important consideration in service delivery.
The ETBs had fewer seats than their diesel counterparts, which pushed the peak load factor higher.

View attachment 45592

Unrelated to some of the other conversations on here, but...
Seeing the 225C at the top of this list is interesting, as there was no MBTA bus service in this area until 2009. (I pulled the 225C out of the data because the T separated the 34E and 70A from their parent routes, but not the 225C for some reason.)

1702599328349.png


Like the 34E or 70A it was was an extended variant of the 225.

1702599927625.png


Just 7 years after adding this MBTA service, there were already an average of 28 riders per trip on the 2 morning inbound trips by the time this bus started the traditional 225 alignment! No wonder the loads were so high by the time it got to Quincy Center.

1702600193490.png


In fact this route was so successful that the T turned it into its own route and changed the terminus to Braintree which now has 20+ trips, even before implementing BNRD.

1702600511418.png


Seeing this go from no service to its own route with 20+ trips a day is a great reminder that simply running MBTA-branded buses on virgin territory can generate surprising ridership, even today.
 
Took a while cause I've been busy but

47
As has already been pointed out the 47 DESPERATELY needs more frequent service. It's quite ridiculous that such an important route that's one of few cross-river services runs every 15-20min at best but drops off to barebones service on weekends.

Starting up at Central Sq during the morning peak, you know when the 47 is coming shortly because the sidewalk becomes difficult to pass through due to how many people are waiting both for a 1 or 47. When the 47 starts boarding it becomes evident that most people are there for it. Departing Central, the bus can be from seats just about full to even eating into standing room when there's a large gap between buses. Down the narrow residential back roads of Cambridgeport there's a steady tossup of a few ons leaving Cambridge with typically the most boardings and a couple of alightings on the Granite Street stops (averaged across the handful of times of taking this route during morning rush but still only a few). *note: on weekends I've seen a few more people on and off in this area during the summer. Crossing the BU Bridge I wish there was some way to incorporate a stop closer to Comm Ave than Lenox St. A small handful will potentially get on here and a few will get off. Alighting passengers primarily walk towards St. Marys rather than backtrack but some do go back towards Comm. I'd imagine the St. Mary's way is the best to go to that section of BU and the BU Central B Line stop. This is what I do when connecting to the B and have seen others who were on the bus along with me turn up waiting for the Green Line (I've only done this transfer on weekends though). At this point when headways are consistent the bus will be using up lower-level standing room but not the upper level. If headways are inconsistent (often the case), things will be cramped and pushed to the upper standing room. A CT2 being caught or catching up also affects the load through the rest of the way to Ruggles. Beacon St is surprisingly light on bus transfer traffic. A handful on and a few off. This might be a combination of the short walk to Brookline for more frequent service to Longwood but also the decent bike infra with how many people bike there. Fenway has more traffic with transfers to the 47 from the D pretty common. Multiple times I've run up the Fenway stairs to catch the 47 (or CT2) with other passengers from the train. On Brookline Ave the 47 can pick up a few and drop off a whole bunch at Simmons, almost empty the standing load at Short St, and drop down to half to 2/3 of the seating at the Galleria. Also at the Galleria is a group of passengers waiting to board and connect to points south like Ruggles and Nubian. The size of these groups varies though because of the overlap with the CT2, 8, and 19. After the swap here the bus will typically sit at around 2/3 seating including a few more boardings and alightings at Boston Children's evening out. The same about even few on/off trend continues down ALP with more boarding common at ALP@Fenway and more alighting at ALP@Longwood. This is again dependent on the duplicate routes. Along Ruggles Street the bus often goes without stopping but if it does it's at Huntington and rarely Annunciation. Ruggles Station drops half the remaining passengers and is typically where I alight myself on weekdays so I can't reliably say how the bus looks the rest of the way. The times I've been on the 47 in this section are a spattering of weekend trips.

Going back towards Central from Ruggles during the evening peak the 47 arrives in the lower busway with anywhere from only a handful to 1/3 of the seats filled. Depending on whether commuter rail trains or the Orange Line let out recently there could be only 4-5 to about 10 passengers waiting to board. The CT2 again plays a role in passenger volume as well since its Longwood to Huntington routing avoids the Ruggles to Park Drive traffic on the Huntington Bus lanes which is especially critical during Fenway gamedays or concert nights. Pickups along Ruggles Street are rare and will only be 2-3 passengers with nobody getting off except for the occasional E transfer which I've done before. ALP@Longwood is the first stop to have more than one passenger boarding but like with the southbound trip the 8, 19, and now CT3 parallel this routing and can take up any traffic trying to get to the LMA or Brookline Ave from Ruggles. Binney Street is where most of the Boarding occurs. The bus goes from at most ~2/3 seating full to cramped standing room Through the upper level. If the standing room doesn't fill here it will be by Short St. My thought is that most LMA workers go to Binney St since it's the first stop in the heart of the LMA and has the highest likelihood of being able to sit. Short Street and Longwood@Brookline are lost causes for space. There are a few people that do alight throughout the LMA. Fenway Station, Beacon St, Carlton St, and Mountfort St all see between 0-5 offs each with Comm Ave the most likely to see greater than one but I've been on a full bus where nobody pulls the stop there on multiple occasions. The Cambridge side of the BU bridge has a few offs on the far side of the rotary there then a handful of passengers trickle off through those residential stops before everyone standing and sitting empties out along Green St. *note: this last section from BU to Central I've only been on during weekday peak a couple times so the alightings in here aren't reliable.

Station Observations
Nubian/Dudley
The time I've spent at Nubian has been primarily later in the evening after 6:30pm

-There are a lot of people that come off of other buses to wait for the 66 here as well as walk here from the surroundings.
-The Silver Line doesn't particularly stand out compared to the 23, and 28.
-When the SL drops off a good amount of passengers either walked away from the station or remain on that same 'platform' which is where the 23/28 then pulls into on the other side.
-The busiest transfers appeared to be between the 23/28 and the Silver Lines/8, and 66. Which makes sense for Roxbury/Mattapan commuters going to the South End or LMA.
As I think @TheRatmeister stated before a lot of this would be eaten up with better Fairmount service combined with better bus connections between Newmarket and the South End.

Forest Hills and Ruggles will go in a separate post cause I'll probably hit the character limit
 
Moving this discussion to the appropriate thread:

@TheRatmeister just pointed out that the Bus Network Redesign (BNRD) Remix Map has proposed frequencies for each route during each time period. This information was used in @Riverside's map here. However, when I last saw it last time, I felt the listed headways were too generic and almost felt like placeholders.

So I took a closer look: I'm in the process of making a spreadsheet with all the proposed BNRD headways for each route.

I have not finished yet, but current progress (Key Bus Routes and regular routes up to 45) already seems to confirm that my earlier impression was, for the most part, correct:

1703123318171.png


There does seem to be a "baseline" standard applied to almost all Frequent Bus Routes, or what I ended up calling "15-8-11-8-11-15" towards the end of this list: 8-min peak, 11-min midday, and 15-min early mornings and late nights.

Only a few routes deviate from this standard, and for the most part, they're what showed up on @Riverside's map: T1, T23, T28, T32, T57, T66, T77 (early morning and late night), T111, T116, and SL5 (merged with SL4). It's interesting to consider both what's included in this list and what's not, but I'll leave it to you all to discuss for the time being.

To the point Riverside made here: it seems like the difference between the "high" and "low" tiers is much less notable than you expected, and for the most part, it's just 11 mins midday vs. 8, 9, or 10 mins. So for the most part, we can actually call the Frequent Bus Routes "12-min routes", instead of the "15-min routes" that the T advertised. Like, the difference between 8-min peak headways and 15-min peaks (implied by the T's branding) is huge - those hours see headways almost halved from the branding.

The "standard service levels" also appear for non-frequent bus routes:

1703123792270.png


I for one don't understand why we're still left with so many "light brown" routes with 50-min midday headways, some of which even have 50-min headways during rush hours. I was in Phoenix earlier this year, and was very surprised by how the vast majority of their buses seem to run every 30 min or better. (They also have quite an interesting network design, for sure.)

But regardless, these numbers do seem to only indicate a planned service level and not the actual schedules. It's very unlikely we'll see all these frequencies play out exactly like this during implementation.

Also, holy F-ing downgrade on the 11.
1703124139575.png
 
Last edited:
Moving this discussion to the appropriate thread:

@TheRatmeister just pointed out that the Bus Network Redesign (BNRD) Remix Map has proposed frequencies for each route during each time period. This information was used in @Riverside's map here. However, when I last saw it last time, I felt the listed headways were too generic and almost felt like placeholders.

So I took a closer look: I'm in the process of making a spreadsheet with all the proposed BNRD headways for each route.

I for one don't understand why we're still left with so many "light brown" routes with 50-min midday headways, some of which even have 50-min headways during rush hours. I was in Phoenix earlier this year, and was very surprised by how the vast majority of their buses seem to run every 30 min or better. (They also have quite an interesting network design, for sure.)
I feel like the answer is that they're questionable at best routes that are politically unpalatable to cut entirely, and are of limited ridership potential + connectivity importance. There may be some merit in a minimal service for a few vulnerable populations, but they don't seem to have significant potential for generating much more ridership with higher service levels. Or at the least - there's other routes that cover the same populations that would see better returns on bumping service levels even higher than these would.

---------

11 - 8th St to Broadway is a 5 minute walk. Unless Broadway is packed so full of buses that it can't hold more, I see no reason why you'd want to add service to this and not the T9

14 - This appears to do basically nothing useful and the sole part of it's route that isn't covered or nearly so by another, better service is....mostly going past dead people in cemeteries, some strip malls, and a golf course. Probably Boston's #1 route for ghost ridership, but not so much for the living.

17 - If you don't want to walk you can just take the T15/T16 that overlap 3/4 of the route. Yeah, there's a transfer for some, but it's between two services with frequent service. And the Fairmount Line to west + Red Line to the seems to really limit how many people are realistically trying to ride this.

30 - makes some sense connectivity wise, but also serves very little directly, it's single-family + some more cemeteries for most of the route. Not sure how big the potential is.

34E - seems consistent with the few other services beyond 128. Within 128 the route is duplicated by the more frequent 34.

38 - I don't see anything about this route that would generate demand. I assume the minimal service is being kept for Faulkner, which also explains the Nubian connection. The southern end has more frequent services to Forest Hills in the 35+36 (which combined, have decent frequency) , where you can get on the Orange Line you're probably heading to anyway. Above Arborway you've got the T39 running frequently or 10min walk to an OL stop

40 - This basically exists to serve one public housing complex, and otherwise just overlaps the 34 (+34E), and Readville also has better options to get you towards Forest Hills.

41 - Appears to exist to connect the VA hospital, BUMC, shelters/homeless services (+ the jail) in a single seat ride. Aside from that specific clientele, I don't really see who'd want to ride this.

44 - ...doesn't look like a very efficient route, and much of the area theoretically has 1 or more frequent routes not very far away.


-----------

tl;dr - If you had one more bus frequency bump to add to the BNRD plans, would you choose to add it to any of these services? I don't think you would.
 
I feel like the answer is that they're questionable at best routes that are politically unpalatable to cut entirely, and are of limited ridership potential + connectivity importance. There may be some merit in a minimal service for a few vulnerable populations, but they don't seem to have significant potential for generating much more ridership with higher service levels. Or at the least - there's other routes that cover the same populations that would see better returns on bumping service levels even higher than these would.

---------

11 - 8th St to Broadway is a 5 minute walk. Unless Broadway is packed so full of buses that it can't hold more, I see no reason why you'd want to add service to this and not the T9

14 - This appears to do basically nothing useful and the sole part of it's route that isn't covered or nearly so by another, better service is....mostly going past dead people in cemeteries, some strip malls, and a golf course. Probably Boston's #1 route for ghost ridership, but not so much for the living.

17 - If you don't want to walk you can just take the T15/T16 that overlap 3/4 of the route. Yeah, there's a transfer for some, but it's between two services with frequent service. And the Fairmount Line to west + Red Line to the seems to really limit how many people are realistically trying to ride this.

30 - makes some sense connectivity wise, but also serves very little directly, it's single-family + some more cemeteries for most of the route. Not sure how big the potential is.

34E - seems consistent with the few other services beyond 128. Within 128 the route is duplicated by the more frequent 34.

38 - I don't see anything about this route that would generate demand. I assume the minimal service is being kept for Faulkner, which also explains the Nubian connection. The southern end has more frequent services to Forest Hills in the 35+36 (which combined, have decent frequency) , where you can get on the Orange Line you're probably heading to anyway. Above Arborway you've got the T39 running frequently or 10min walk to an OL stop

40 - This basically exists to serve one public housing complex, and otherwise just overlaps the 34 (+34E), and Readville also has better options to get you towards Forest Hills.

41 - Appears to exist to connect the VA hospital, BUMC, shelters/homeless services (+ the jail) in a single seat ride. Aside from that specific clientele, I don't really see who'd want to ride this.

44 - ...doesn't look like a very efficient route, and much of the area theoretically has 1 or more frequent routes not very far away.


-----------

tl;dr - If you had one more bus frequency bump to add to the BNRD plans, would you choose to add it to any of these services? I don't think you would.
@HenryAlan i think commented a while back on the 14. Not sure what populations it serves but I just think it’s kinda cool I can walk from my house in rozzie and get a one seat ride to the JP VA. Come to think of it, it probably does exist for the VA—I bet you there are a few lines that seem wonky but exist to ensure vets can get to the hospitals without multiple changes. Also, I think this is the only bus route that gets you from Roslindale Sq to Dorchester/Franklin Park. Otherwise you’d have to change at FH. And I do think a Rozzie<~>Dot connector makes sense.

38 — somewhere on here, at some point in 2023, I asked about this line and got some educating from a bunch of folks on here. You’ll have to find where it is, but there was, if I remember correctly, some good discussion that explained why this line exists.
 
@HenryAlan i think commented a while back on the 14. Not sure what populations it serves but I just think it’s kinda cool I can walk from my house in rozzie and get a one seat ride to the JP VA. Come to think of it, it probably does exist for the VA—I bet you there are a few lines that seem wonky but exist to ensure vets can get to the hospitals without multiple changes. Also, I think this is the only bus route that gets you from Roslindale Sq to Dorchester/Franklin Park. Otherwise you’d have to change at FH. And I do think a Rozzie<~>Dot connector makes sense.

38 — somewhere on here, at some point in 2023, I asked about this line and got some educating from a bunch of folks on here. You’ll have to find where it is, but there was, if I remember correctly, some good discussion that explained why this line exists.
Yeah, I'm a fan of the 14 for a few reasons.
  1. It is the only direct connection between Roslindale and the heart of Roxbury
  2. It has a very convenient transfer to several Dorchester serving lines
  3. It is pretty much the only service through the American Legion Hwy corridor, which is fairly densely settled
  4. I'm not sure I'd take it from Rozzie Square all the way to Heath St/VA Hospital, but I bet there are people in Roxbury who find that to be a convenient connection.
Basically, if you got rid of the 14, you'd need to recreate it by rerouting some combination of other buses.

As for the 38, I somewhat recall that discussion, but not the specifics. It's obviously pretty crucial for a bus to serve the Faulkner. I think there is also a decent amount of ridership from people who live along the section of Centre St. between VFW Parkway and Holy Name rotary. The bus network redesign will hurt both groups, in that the 38 will no longer provide as efficient a connection to the Orange Line.
 
Yeah, I'm a fan of the 14 for a few reasons.
  1. It is the only direct connection between Roslindale and the heart of Roxbury
  2. It has a very convenient transfer to several Dorchester serving lines
  3. It is pretty much the only service through the American Legion Hwy corridor, which is fairly densely settled
  4. I'm not sure I'd take it from Rozzie Square all the way to Heath St/VA Hospital, but I bet there are people in Roxbury who find that to be a convenient connection.
Basically, if you got rid of the 14, you'd need to recreate it by rerouting some combination of other buses.

As for the 38, I somewhat recall that discussion, but not the specifics. It's obviously pretty crucial for a bus to serve the Faulkner. I think there is also a decent amount of ridership from people who live along the section of Centre St. between VFW Parkway and Holy Name rotary. The bus network redesign will hurt both groups, in that the 38 will no longer provide as efficient a connection to the Orange Line.
Good points.

And while I forget all the deets on the 38, it does serve the Bellevue residential hood, and it also serves the Faulkner a you pointed out and that alone makes it worth preserving. I rather wish the Centre St corridor beyond VFW had more of a transit need, since that stretch would do very nicely with a BRT lane given how unnecessarily wide it is
 
I've taken the 14 quite a few times to get between Heath and the OL/Jackson. It regularly came into Jackson with 5-7 people on board. Most of them alight at Jackson with only a couple remaining on. I was usually not the only person boarding towards Heath.
Coming the other way in the mornings there was typically 3-5 people including myself that board at the E terminus and it'd pick up a few more on the way to Jackson. Never went beyond to ride the whole thing though.
It appeared as though it was a few people's lifeline to get to the VA but its lack of frequency and unreliability due to traffic on Heath, especially during the school rush, make it so only those who must use it do. The stretch along Heath St is pretty densely populated too with some new housing. Could probably benefit quite a bit from increased service.
 
  • Like
Reactions: FK4
According to their project page, the Summer St pilot bus/truck lane program started on Monday, December 4.
I rode the 7 bus through the Summer St pilot bus lane today. My impressions may not be accurate, as I couldn't see everything clearly on a bus that was quickly moving, but... Either they haven't finished the painting, or it fell way short of initial plans.

For most stretches of the road, the bus lanes were not painted red when not nearing junctions. Only when entering the stretch (just past a junction) do you see a short distance with red paint and "BUS ONLY" (I'm not even sure if every junction said that, and IIRC many junctions didn't). When exiting the stretch (approaching a junction), there are alternating red and white stripes with the right turn sign, but most if not all junctions don't say "BUS ONLY" before that.

Yeah, my recollection is very vague here, but I think there will be many drivers who have no idea that's a bus lane.

While most cars in my direction didn't use the bus lane, there were already one or two who took the bus lane and didn't seem to be turning right.
 
Interesting about the paint pattern. That's similar to how some of the rush hour only bus lanes have been painted (eg Washington St. in Roslindale). I get the concept for a lane that is only active for a few hours, but I think 24 hour bus lanes should be 100% painted (and separated by flex posts).
 
From the article:
While no MBTA bus routes currently use Rutherford Avenue end-to-end, the busway would benefit several existing and planned routes, including the 111 and the 109.

In the longer term, as new buildings bring more riders into the neighborhood and nearby areas, like lower Broadway in Everett, the City of Boston hopes to create a new bus rapid transit route that Sullivan Square and downtown Boston.

The MBTA's ongoing Silver Line Extension study identified Rutherford Avenue as a likely link for potential new bus rapid transit lines between Everett and Kendall Square or downtown Boston.

“It is difficult for the MBTA to make service commitments or do planning more than a few years out for what kinds of new services might exist. When we think about a design for this project, we’re thinking 50 years into the future," said Franklin-Hodge. "What kind of transportation needs will exist then?"
It's curious that the SL6 study pretty heavily favored Kendall over Haymarket (in both ridership projections and community feedback), and even among the Kendall alignments, it favored Washington St and McGrath Hwy (Alt 4) over Rutherford Ave and Gilmore Bridge (Alt 5). Could it be that at the time, they weren't able to rely on bus lanes from the Rutherford Ave redesign as a certainty, and/or residents still expected buses to be stuck in traffic?

While I do think the Rutherford bus lanes are important in the long term for an Everett-downtown BRT connection, they might only be effective when both the Rutherford Ave bus lanes and the downtown bus priority corridor are completed, allowing an OSR beyond Haymarket.

I do wish they can add a stop for the 111 near the Galvin Memorial Park area. This would improve service to Charlestown pretty dramatically without expensive Tobin modifications.
 
While I do think the Rutherford bus lanes are important in the long term for an Everett-downtown BRT connection, they might only be effective when both the Rutherford Ave bus lanes and the downtown bus priority corridor are completed, allowing an OSR beyond Haymarket.
OSR? Google is failing me.
 

Back
Top