Dwell times

Last November, WBZ-TV interviewed Jeffrey Gonneville, the Deputy General Manager of the MBTA. The interview took place while touring the MBTA’s assembly factory in Springfield. One of Mr. Gonneville’s comments was, “We really want to hold to about a forty-five second dwell time at all of our ‘T’ downtown stations.”

I’m going to assume that “dwell time” is the amount of time that a train is stopped in a station. My one-word response to Mr. Gonneville: “Unacceptable”.

If one goes to YouTube.com, one can find a plethora of subway videos. I found seven videos taken out the front window of trains in New York City. I found four #7 express train videos, one “A” train, one “C” train, and one “J” train video. Each video was uncut for an entire run of the train. One #7 train had an average dwell time of 41.8 seconds. The remaining six trains had an average dwell time of 30 seconds or less including the “J” train which had the best dwell time average of 20.5 seconds. The average for all seven train runs was 28.7 seconds.

There used to be a sign at Harvard that said, “8 MINUTES TO PARK ST.” The scheduled time is now 11 minutes, 37% longer. There used to be a sign at Blandford Street that said, “9 Minutes TO PARK ST.” Today’s schedule time is 13 minutes, 44% longer. When Riverside opened, the scheduled time to Park Street was 35 minutes. Today’s schedule time is 47 minutes, 35% longer. Ah, progress.

Now that the design of the cars is complete and manufacturing well underway, is it really useful to discuss dwell time in the factory?

Dwell time is largely affected by passenger loading. A car packed to the gills will have a much longer dwell time. The discussion needs to be moved back to downtown Boston, to headquarters or a board room, to center on train headways.

Do the new trains' door controls have a provision to reclose a door that sprang back open due to someone going in or out at the last second, without re-opening all of the other doors only to have a straggler repeat the process at another door?

Perhaps we need to address other operational delay as well as dwell time within a station.

(copied from another post) The signaling and blocks may need to be revamped to allow closer train spacing. Let's imagine (Red line pile up as the example) an inbound train stopped, open-doored, at Kendall and the next train stopped, open doored, at Central. When the first train finally clears out of Kendall, then the second then just starts up from Central to Kendall and the third is back at Harvard. Whereas if additional trains wait, closed-doored of course, between stations, the second train only has to go the other half of the way to Kendall and the third train can proceed out of Central much sooner with a fourth inbound train already halfway from Harvard to Central.

A sophisticated system would not allow trains to get that close together unless they were behind their original scheduled times.

Back in the 1960's I recall the trains, including the PCCs on the Green Line, "roaring" into the stations. I would guess that they hit a straightaway station entrance at no less than 25 MPH under clear track conditions. PCC streetcars were designed for rapid acceleration, too. Today I have heard of a 10 MPH speed limit within stations. This also affects the overall trip time.

Yellowspoon wrote: ↑Mon Feb 10, 2020 10:26 am There used to be a sign at Harvard that said, “8 MINUTES TO PARK ST.” The scheduled time is now 11 minutes, 37% longer. There used to be a sign at Blandford Street that said, “9 Minutes TO PARK ST.” Today’s schedule time is 13 minutes, 44% longer. When Riverside opened, the scheduled time to Park Street was 35 minutes. Today’s schedule time is 47 minutes, 35% longer. Ah, progress.

Just because they had a big sign that said "8 minutes to park street" doesn't mean that was actually the case. Dwell times are going to go up as passenger numbers go up. 45 seconds of dwell during off-peak time is atrocious but during rush hour at a core downtown station is pretty understandable given the volume of passengers.

Disney Guy wrote: ↑Tue Feb 11, 2020 11:06 amA sophisticated system would not allow trains to get that close together unless they were behind their original scheduled times.

I can't imagine the T implementing a system where train spacing is a factor of how late a train is.

It seems to me that the solution to dwell time is to run more frequently. While visiting Moscow in 1975, I remember seeing a clock on the wall in a Metro station; the clock would count down from two minutes, and when it got to zero, a train pulled in. It would sit there for the prescribed number of seconds; then a recorded voice said "осторожно, двери закрываютсья" ("be careful, doors are closing"), and then the doors closed and the train pulled out.

However inept the Soviets may have been in other areas, they did know how to run trains.

When the "N Minutes to Park St." signs were put up (by my great-grandfather's company), they were accurate. I believe they were still accurate in the mid-70s when I returned to Boston as a student. But speed limits have been reduced: I used to see 55 MPH routinely on clear track between the curve onto Main St. and Kendall Sq. The Green Line got slower with each accident, and what used to be speed-timed signals approaching restrictions turned into stop & proceed. The bargain basement heavy rail signal implementations of the 1970s and 1980s added insult to injury: block boundaries in the middle of station platforms (Community College and Charles/MGH inbound, among others) led to stopped trains being unable to open their doors. Another blow was reduced train crews. 4-car trains with two guards each minding two cars worth of doors had much less trouble getting them closed than one controlling six cars. And after an Orange line train comes to a stop, we all stand at the door while the operator gets up (and at some stations steps across the cab), opens the window and checks (per the rules) before opening the doors. Then another delay reversing that on departure.

The coup de grace has been the total failure of the Commonwealth/MBTA to add capacity. Additional capacity from 6-car trains mostly went to undoing the damage from lower speeds and longer signal blocks. The Boylston St. tunnel has been at capacity during rush hour for at least 2 decades, maybe 3. But there's no money to dig new tunnels so the Silver Line to Roxbury is just diesel buses fighting traffic in a special paint job. And the Urban Ring was perceived as threatening Financial District property values, so it got demoted to more buses and then ignored. So everything is totally packed in the busy direction every weekday rush hour, and dwell times soar. Longer Green Line cars will add a bit (at cost of major work on the Lechmere Viaduct) but that can't be repeated without major work on the tunnels.

We really need to get like buttons on here somehow! The above is seconded.

The solution is to connect the D and E branches and a new tunnel connecting Back Bay to the old Pleasant st Tunnel. Future-proofs the GL for 50yrs. Not cheap but has to happen.

Connecting the D & E line? How would that even work. Reopening the Tremont St portal seems like a great idea, but that won't reduce dwell time on the Central Subway. More capacity, station bypass tracks, better timing/signals, platform screen doors are the kind of things needed.

Street running down Route 9 for two blocks or so from the corner of Huntington/South Huntington to a junction on the curve just west of Brookline Village. You'd have to acquire and demolish a few buildings at the latter location.

The MBTA's dwell times are unconscionable. From @FriendCristoph on twitter:



100 second median dwells at DTX is significantly worse than what we do here in NYC (60s is considered really high), and NYC has much higher peak hour loads than does Boston...and we really are not good at dwell control! As @jvbv notes, signal system design aggravates this issue, but there's only so much you can extract with signal mods when your operating fundamentals are as bad as this. Some suggestions:

- Buy rolling stock with CCTV in cab so operators don't have to get up to operate doors
- Allow more aggressive door control/enable local recycle functionality
- Consider rolling stock with more doors (MBTA seems to have gone for wider doors on the new OL cars, which is good to see)
- Get better at dispatching even service to smooth loads across trains
- Pay attention to your average as well as your outliers. 50s dwell at medium-to-lightly used stations even in NYC is almost unheard of. Running a tight railroad at those locations as well as at the Parks and DTXs of the world allows you to cut runtime, permitting increased equipment utilization, more trains, and thus lower dwells, etc etc etc.

And just to give some international context here: Moscow Metro runs 40tph (requiring max. 30-40s dwells) on fixed blocks with passengers loads much greater than anything the MBTA faces. A bad station in London (smaller rolling stock, higher loads than MBTA) has 50s dwells.

The MBTA has it easy, but doesn't do well.