General MBTA Topics (Multi Modal, Budget, MassDOT)

Regarding lifting slow zones on the Green Line central subway tunnel post-monthlong shutdown:

I had a chance to ride the Green Line between North Station and Longwood Medical Area westbound, and eastbound back to Haymarket.

The slow zones between North Station and Gov't Ctr are towards Haymarket. I saw an S6 sign before curving into Haymarket, and then I saw 2 "RESUME SPEED" signs in quick sucession past Haymarket. The second "RESUME SPEED" sign appeared to be shortly before the merge from the Gov't Ctr loop. So it seems the 5 MPH WB speed restriction between Haymarket and Government Center doesn't affect Boylston St. streetcars looping to return to Park St.

The trip from North Station to Longwood Medica Area was extremely slow. The streetcar left North Station at 13:47 (1:47 p.m.), and did not get into Longwood Medical Area until 14:18 (2:18 p.m.), a 31 minute trip that used to take 20 - 22 minutes back in May 2018.

Much of the central subway tunnel still seems to be capped at 15 MPH (25kmh) from Government Center to Symphony. There may have been some S25 (40kmh) speed limit signs somewhere near Gov't Ctr, or Arlington, but I couldn't see them or tell them apart from the S15 signs in the dimly lit dark tunnel. On curves and switches in the tunnel between Government Center and Symphony, they are limited to 6 MPH (9.6 kmh) with the speed limit signs stating as such.

At Northeastern there is a 10 MPH (16 kmh) limit at the switches, then after that it is S25 with no additional speed limit signs before Longwood Medical Area.

Heading eastbound there was too much rush hour crowding so I could not observe any speed limit signs, the trip was much faster and I got to Haymarket from Longwood Medical Area in about 23 minutes (historically around 18 minutes in May 2018).
North Station to Government Center is still slow, with only a brief 15 mi/h zone before the 6 you reference; I think Haymarket to Govy is 6 the entire way right now; Govy to Park has just been raised to 20; there are one or two 25's between Boylston and Copley; the switch at Arlington we can do 10 over. I haven't driven out on Huntington Ave since training because of the diversions, but I have Medford work tomorrow so I'll report back on what I see if I end up being the driver for a trip.

Some of the signage in the tunnel leaves a lot to be desired.

Your inconsistent experience is consistent with how things work over here on the Green Line. I wonder if you got one of my classmates as your driver or if there was just a lot of congestion or some other problem.
 
Apparently, the most recent Green Line shutdown did little to improve travel times in practice, according to this tweet shared in the Reddit post below. But as the comments pointed out, it demonstrates the unreliability of the Green Line (both operations and vehicles) more than the state of track repairs.


(Original tweet)
 
I haven't had a chance to ask this here, but several weeks ago as I was walking into Costco in Everett, I turned around to hold my daughter's hand I noticed the Downeaster flying through the line next to Assembly... how often does something like that happen, because that was the first time that I have seen that. I'm pretty sure that would have been a sight having seen that fly through Malden Center Station and Oak Grove.
 
I haven't had a chance to ask this here, but several weeks ago as I was walking into Costco in Everett, I turned around to hold my daughter's hand I noticed the Downeaster flying through the line next to Assembly... how often does something like that happen, because that was the first time that I have seen that. I'm pretty sure that would have been a sight having seen that fly through Malden Center Station and Oak Grove.
Once in a while the Downeaster takes a detour along the haverhill/reading line instead of crossing over to the Lowell line on the wildcat branch. I’m not sure why the train you’re referring to was detouring, but it’s usually because of scheduled track work that creates constraints for the T. I’ve ridden it along that line once, and it was a pretty cool route to take!
 
MBTA vacancies update for buses, January 2024:

View attachment 46705

Here is an update regarding the MBTA bus service/operator shortage and the MBTA's efforts to resolve this issue as of February 2024:

DateCountVacantChange% of pre-COVID maxMax count
5/2/20231,622201N/A88.97%1,823
6/29/20231,611212-1188.37%1,823
7/26/20231,548368 (275 pre COVID)-63 (-93)84.91% (80.79% BNRD)1,916
8/30/20231,559357 (264 pre COVID)+1185.51% (81.36% BNRD)1,916
10/4/20231,572295 (251 pre COVID)+62 (+49)86.23% (84.19% BNRD)1,867**
10/23/20231,644223 (179 pre COVID)+7290.18% (88.05% BNRD)1,867**
11/28/20231,642225 (181 pre COVID)-290.07% (87.94% BNRD)1,867**
12/29/20231,635***232 (188 pre COVID)***-7***89.68% (87.57% BNRD)****1,867**
1/25/20241,697170 (126 pre COVID)+6293.08% (90.89% BNRD)****1,867**

****The formula the MBTA used to calculate bus vacancies has been changed again this January, to exclude in trainees from the operational count (those in training will now be included in vacancies). No changes to historical data was provided by the MBTA, meaning the data since December 29, 2023 is not directly compariable with May - November data.

**Updated figure from the Boston Globe

1707408040835.png
 
Last edited:
Once in a while the Downeaster takes a detour along the haverhill/reading line instead of crossing over to the Lowell line on the wildcat branch. I’m not sure why the train you’re referring to was detouring, but it’s usually because of scheduled track work that creates constraints for the T. I’ve ridden it along that line once, and it was a pretty cool route to take!
Related thought: In a world where Reading is converted to OLX, would this be an issue for Downeaster operations if they don't have a backup route?
 
Related thought: In a world where Reading is converted to OLX, would this be an issue for Downeaster operations if they don't have a backup route?
Not really. The diversions just give the T the luxury of daytime instead of nighttime maintenance windows. It's nothing essential. The rest of the DE north of Wilmington Jct. has no backup route, and it's not cancellation-prone.
 
Apparently, the most recent Green Line shutdown did little to improve travel times in practice, according to this tweet shared in the Reddit post below. But as the comments pointed out, it demonstrates the unreliability of the Green Line (both operations and vehicles) more than the state of track repairs.


(Original tweet)

According to transitmatters' data, the Green Line is significantly slower over the North Station <> Kenmore stretch than it was a year ago. WHY?
That is not a rhetorical question. Any and all answers welcome.

For perspective, that stretch has undergone a total of 30 days of shutdowns for track work under Eng's administration. I am not blaming him (that should go without saying), but with how successful the shutdowns were for the Ashmont Branch, why were these an utter failure? I'd understand getting back to early 2023 travel times and hitting a wall due to other issues (operations and vehicles have been cited), but I can not come up with a good answer as to why it's slower than one year ago.
 
According to transitmatters' data, the Green Line is significantly slower over the North Station <> Kenmore stretch than it was a year ago. WHY?
That is not a rhetorical question. Any and all answers welcome.

For perspective, that stretch has undergone a total of 30 days of shutdowns for track work under Eng's administration. I am not blaming him (that should go without saying), but with how successful the shutdowns were for the Ashmont Branch, why were these an utter failure? I'd understand getting back to early 2023 travel times and hitting a wall due to other issues (operations and vehicles have been cited), but I can not come up with a good answer as to why it's slower than one year ago.
There are still slow zones between Copley and Kenmore.
 
I have noticed the Green Line trunk Between Haymarket and Hynes Convention Ctr & Brigham Circle has failed to deliver promised results, with actual data only showing a small fraction of the promised travel time savings.

(The reason I select origin destinations heading to Haymarket is primarily since I need the data to evaluate the amount of travel time to reach the Haymarket Congress St. busway for the bus connection, hence Haymarket is selected as the destination).

To figure out what's going on I had to download each month of data from TransitMatters as the data dashboard regularly refuses to load more than 3 - 9 weeks of data for the Green Line trunk at one time, so I downloaded all data to 2016 to find the slow and fast points for the Green Line trunk.

I then took the data from TransitMatters and converted it into 7 day rolling averages, in order to smoothen out the data and get better trends by reducing the effect of outlier data and weekend travel differences compared to weekdays.

For the stretch from Kenmore to Haymarket eastbound, I have got this data. The Green Line trunk is slower than the pre-March 2023 travel times, although it is not as slow as it was around April-May 2023. It is still over 1.5 - 2 minutes slower than Spring 2018, which had the fastest travel times between Kenmore and Haymarket eastbound outside of COVID. Travel times have only decreased by around 2 minutes compared to the peak in October 2023.

Note: There are a few slow zones from Kenmore to just east of Hynes Convention Center.

1707425526652.png


For the stretch from Longwood Medical Area to Haymarket eastbound, travel times are slightly slower then they were in Fall 2022 or Winter 2023. It is also around 1.5 minutes slower than it was in Spring 2018, the fastest travel times outside of COVID. There are zero slow zones Eastbound (EB) from Longwood Medical Area to Haymarket. Travel times have decreased by around 1.5 minutes compared to the peak in October 2023, although peaks in October 2019 and June 2016 are higher (slower) than the slowest travel times in October 2023.
1707425642461.png
 
Last edited:
According to transitmatters' data, the Green Line is significantly slower over the North Station <> Kenmore stretch than it was a year ago. WHY?
That is not a rhetorical question. Any and all answers welcome.

For perspective, that stretch has undergone a total of 30 days of shutdowns for track work under Eng's administration. I am not blaming him (that should go without saying), but with how successful the shutdowns were for the Ashmont Branch, why were these an utter failure? I'd understand getting back to early 2023 travel times and hitting a wall due to other issues (operations and vehicles have been cited), but I can not come up with a good answer as to why it's slower than one year ago.
My first guess was due to reliability (e.g. trains bunching), but headways from Boylston to Copley seem to have improved by around 10 seconds from 2023 to 2024 for the week of Feb 1-7: (I only used a week because doing anything more than that in 2023 gave me an error in TransitMatters dashboard)
WeekAvgMedian
2/2/23 - 2/8/23 (Thu-Wed)108s107s
2/1/24 - 2/7/24 (Thu-Wed)97s99s

However, looking at headways at Boylston westbound on a single day, it indeed suggests that while Green Line trains are running more frequently on average than a year ago, westbound trains are running less reliably.
DateMeanStandard Deviation
Tue 2/7/2023141.2s105.4s
Wed 2/8/2023147.5s106.4s
Tue 2/6/2024152.8s125.6s
Wed 2/7/2024142.9s120.6s

(Northbound trains are less reliable too, though on a few days I checked, average headways are actually worse than 2023, even though the weeklong average is better.)

As for its cause or how bad that is, I don't know. I did try to make a histogram, but not sure what I'm getting from it.

GL Headway Histogram Feb 2024.png


So... Yeah, this is kind of an analysis that went nowhere lol. Still unclear is this was the actual reason for the lack of travel time improvements.
 
To figure out what's going on I had to download each month of data from TransitMatters as the data dashboard regularly refuses to load more than 3 - 9 weeks of data for the Green Line trunk at one time, so I downloaded all data to 2016 to find the slow and fast points for the Green Line trunk.
Question: How did you download the data? Is there a TransitMatters API or something?

As for the results: What happened in October 2019 that caused the huge spike in travel times? Did slow zones already exist back then, or is it just seasonal variations due to dwell times? It appears that we're on par with Jan 2020 (low point during the Oct '19 -- Covid period) and Oct 2018 (another spike before 2019) now.

Perhaps subtracting dwell times from the data (which TransitMatters also provides) may be more representative of track speed.
 
Last edited:
Apparently, the most recent Green Line shutdown did little to improve travel times in practice, according to this tweet shared in the Reddit post below. But as the comments pointed out, it demonstrates the unreliability of the Green Line (both operations and vehicles) more than the state of track repairs.


(Original tweet)
As an operator, I can attest to this. Ops and vehicles have an outsized role in dragging the Green Line. Also, speeding is definitely a thing; on the heavy rail lines, all runs get faster because the signal system prevents you from speeding. On the Green Line, only the runs of people who follow the rules get faster.
 
Question: How did you download the data? Is there a TransitMatters API or something?

As for the results: What happened in October 2019 that caused the huge spike in travel times? Did slow zones already exist back then, or is it just seasonal variations due to dwell times? It appears that we're on par with Jan 2020 (low point during the Oct '19 -- Covid period) and Oct 2018 (another spike before 2019) now.

Perhaps subtracting dwell times from the data (which TransitMatters also provides) may be more representative of track speed.
Manually doing each date range and figuring out the max range before the data dashboard errors out, then going through each range one by one until I had all the ranges. (The max range seems to change from time to time, right now it seems to allow only up to 3 weeks max and anything more errors out atm. When I downloaded the whole data a few days ago the data dashboard allowed me up to 8 months before it errored out).
 
As an operator, I can attest to this. Ops and vehicles have an outsized role in dragging the Green Line. Also, speeding is definitely a thing; on the heavy rail lines, all runs get faster because the signal system prevents you from speeding. On the Green Line, only the runs of people who follow the rules get faster.

Right and this would explain why there aren’t more time savings, but doesn’t explain why it was quicker in years past.
 
Last edited:
Dispatcher shortage? Green Line tends to wait at signals a lot.
I'm thinking it's dispatching/signals. The amount of times and length of time seemingly sitting still for no reason has been much more noticeable recently. But as I said upthread, I have nothing to back this up except for anecdotal experience...

Maybe our operator can give some insight?
 
Right and this would explain why there aren’t more time savings, but doesn’t explain why it was quicker in years past.
This is definitely true.
I'm thinking it's dispatching/signals. The amount of times and length of time seemingly sitting still for no reason has been much more noticeable recently. But as I said upthread, I have nothing to back this up except for anecdotal experience...

Maybe our operator can give some insight?
I'm still quite new (hired at the end of October, revenue service mid-January) at the job but have plenty of experience riding, albeit I have always preferred riding Orange over Green through downtown if there isn't a major headway gap on the Orange.

I don't have any hard evidence either but as a rider I'll agree that I have noticed an increase in the number of standbys on the Green Line. As both a rider and operator, I am inclined to blame dispatch.

Why dispatch does the things they do, I don't know.... during the last diversion I got held at St. Mary's for 6+ minutes multiple times. Having to explain that one to riders was fun. They also expressed me from Reservoir to Riverside a few days ago but they didn't bother calling me until I was already off the platform at the Res, so they had to make it express from Newton Centre (expressing from Chestnut Hill would have been a violation of policy). This came just minutes after they caught me right as I was opening my doors at Longwood to go express to the Res. I am fully convinced they need an extra dispatcher or two. One time in training we waited over 10 minutes to officially have access to the right-of-way just so my instructor and I could throw the switches at Chestnut Hill Ave going from Cleveland Circle to BC, all while our train was blocking traffic. Obviously the rule is FTA-mandated, but the dispatchers were just ignoring us and letting us sit there even though there was a gap in service on the BC line which would have been a good time to slide right onto the line. They, in my opinion, need more people on light rail dispatch.

As others have mentioned, the signal system is ancient and its ability to have high frequencies is counteracted by its ability to let trains crash. There is a signal coming into Park Street westbound on the D/E platform that stays red even when the platform is empty, then changes after several seconds. There are also a few "time light" signals (the most annoying of which are near Kenmore and Hynes) that serve to slow us down on purpose. They aren't even controlled by OCC or anything, they're just always defaulted to red and only change when the train has been in the current block for long enough (provided the block ahead is clear). They change directly to green, unless a train is close enough to warrant a yellow. These aren't new, but I don't know if the one at Park might have been recently re-programmed to also serve as a time light or if it's just faulty. I have no hard evidence for this whatsoever but I wonder if the signal system has been tweaked since the FTA came in -- it wouldn't surprise me at all.

The Green has always been the line with the longest and most frequent standbys in my anecdotal experience; I've been on trains that have been held on the heavy rail plenty of times but aside from what I refer to the "standby of fate" on the Red Line (11-minute standby at JFK that allowed me to say goodbye to my dying grandfather because the standby made me miss the Orange Line and thus a train at North Station), the only 10+ minute standbys for no obvious reason I have seen have been on the Green.

I should also note rolling stock is an issue on the Green Line. Maybe I already have. I got a bad train for my last trip Tuesday out of Medford, I was driving the 8 in front and the trailing 7 kept springing propulsion faults that dramatically slowed the trip down. I would reset the propulsion at every stop and it would just come back. It was bad enough that I was giving the train full power coming into Govy on the westbound and could only get the train to 5 mi/h (in a 6 so that's okay, but with the master controller to full throttle is not normal), and then coming back (because the officials and dispatch had me keep moving rather than swap me with a train headed to Riverside/Reservoir somewhere) I almost stalled on the incline coming toward Science Park. The train was slowing down, already below 5, again giving it full throttle. Stopped my train, reset the propulsion (which is as simple as clicking a button), and was able to kick it to 12 or 13 before the fault came back but I was able to give myself enough momentum to get up the hill. I then got expressed from Lechmere to Medford/Tufts. Anyway, one train with a particularly bad propulsion fault can mess up the subway (and if said train goes onto Riverside or Medford.... good luck! I've had bad trains like this on Riverside multiple times and it's awful). I feel like the wire car has also been out a lot lately, which tends to goof it up too.
 
The habit of the GLX expressing to Medford/Tufts once it’s on the downslope into East Somerville, rather than at Lechmere is annoying. Usually the train has just loaded a bunch of folks waiting on the platform, only to disgorge them one stop later. It seems more efficient to leave them at Lechmere.
 

Back
Top