Muni to Test Double-Train Loading in Metro Stations This July

Photo: ##

Muni will test out a method to improve metro operations this July, letting two trains load passengers on the same platform simultaneously, according to SFMTA spokesperson Paul Rose.

Few details on the test run are available, but Muni officials have discussed making the change for several years as a way to reduce back-ups in the subway underneath Market Street. Currently, multiple trains often sit at the same platform, but only the one at the front can load. Transit officials have said that allowing two trains to load, one behind the other — also known as “double berthing” — would require changes to the Muni Metro’s automatic train control system, new training for rail operators, and an upgraded digital sign and announcement system.

Mario Tanev of the SF Transit Riders Union said the organization hasn’t taken an official position on the change, but that it supports Muni’s efforts to try out new solutions to improve service. “Congestion in the subway is a serious problem,” he said. “There needs to be solutions.”

  • Admittedly a transit nerd

    About time! So excited!

  • Any ETA of when those new destination signs are coming thru? July perhaps?

    Also – I would hope the recently-refurbished automatic train control system would be able to handle this change in operation.

  • Nice! But I hope they test this long enough for riders to get over the initial confusion.

  • Anonymous

    Exploring what it would take to allow double-berthing is long overdue; it seems like such a simple way to speed up Metro service.

  • Yes! I always thought it so strange that the platforms so incredibly long, but yet you have to be at the front of the line for the doors to open.

  • ywhynot

    Wow! Satan must have his uggs and pea-coat on right now!

  • Henry

    From what I understand, double-berthing will only occur at Civic Center, Powell, and Montgomery Stations, right?

  • They used to do this around 15 years ago. My question is why did they ever stop?

  • Anonymous

    I don’t think they need new signs. There should be enough there already to make it work. I hope they were just referring to software upgrades.

  • Anonymous

    Don’t forget Embarcadero! That’s where it’s needed most.

  • Anonymous

    Confusion? I imagine there may be some scrambling to get into the proper position, but when trains arrive, those blinking letters on the sign should attract a lot of attention! (I hope!)

  • That assumes you can even see the blinking letters. From where most people wait at Powell, for example, it would be damn hard to see that a second train was boarding.

  • Anonymous

    And after this… how about 3 car N-Judah’s or at least S-Shuttle’s for Ballpark service!

  • For it to work effectively there should be designated boarding zones for certain lines (like boarding area A for J and N and area B for K, L, and M.) That way people don’t need to always check the sign for boarding location or having to rush to the location when the train comes, which is a problem for disabled riders.

    Of course it would make double berthing less effective but still better than the current set up.

    The Downtown Seattle tunnel has this set up for buses (light rail only stop when there are no buses and use the whole platform).

  • Ted King

    Something to remember – SFMuni Metro’s platforms were designed for BART trains. When they gave them to SFMuni they made a HUGE mistake by NOT lowering them to a low-floor standard. That gave us the headaches of dual-mode stairs and expensive ramps and platforms outside of the tunnels.

  • Mario Tanev

    One possible solution to the confusion that could be caused by double-berthing is to make the rear train offload-only, and have it stop again to board passengers (some will cunningly run to board, though, and others will freak out, afraid of missing their train, so confusion won’t be completely eliminated).

    It would at the very least, however, solve the frustrating experience of waiting for minutes on end just to exit the stopped train.

  • Mario Tanev

    Frankly, I don’t see what 3 car N trains would solve. If Muni has a shortage of operational train cars (which it does), 3 car trains would simply mean lower frequency, and thus won’t solve overcrowding and will generally worsen the experience (although, labor may be cheaper). Longer trains however may make sense once the frequency is very high already, and smaller trains tend to interfere with each other – so I can see that working on a shuttle-like service in the subway.

  • Jamison Wieser

    Most of the infrastructure outside of the subway can only accommodate a 2-car train and at the loops where they turn back, just a pair 3-car train would block two intersections.

    Because of this, the trains were only ever designed to pull a third train in a tow mode. Since there was never any expectation to run 3-car trains it wasn’t worth the time or money to add and add the functionality. It’s a software issue: the couple carries breaking and acceleration to the third car in two-mode and it’s completely doable.

    Double-birthing is an option that comes along with the new announcement system, are part of tunnel control system upgrades underway. Operators, controls, and managers will need retraining anyway, so a lot of these costs are fixed whether we try it or not. But there is an opportunity cost if it does prove more effective.

    Like Mario explained, It’s really a matter of efficiency: a train can run faster, then in a given period of time it can make more runs. Flipping that equation around, that also means the current frequency (or headway) can be maintained with fewer trains.

    That’s where you get the labor cost saving. If we can get them to run efficiently (that doesn’t just mean faster, being predictable counts) that frees up cars for more 2-car trains. And that increases capacity. and so on…

  • Spoken like someone who has never heard “2 car N in 2 minutes. 2 car N in 4 minutes. 2 car N in 6 Minutes. 1 Car J in 7 minutes”

  • Anonymous

    The underground system was originally designed to accommodate 3- or 4-car trains, even with different lines on the same train … you could have a KLM train or an NNJ run through the system, theoretically, and the cars could decouple above ground.

  • Anonymous

    I think they used to have multiple-line trains, but the Breda cars weren’t designed for such frequent decoupling.

  • Anonymous

    Low-floor boarding in a subway is not a good idea. Elevated platforms keep passengers moving.

    Bart trains were never designed to run in Muni tunnels, or vice-versa. One is light rail, the other heavy.

  • Anonymous

    Good point! It would e nice to see the other waiting area put to good use. The main one gets overcrowded sometimes.

  • Another question: How will the voice announcements (aka Ms. Muni) work with this?

  • Anonymous

    “Approaching: 1-car J and 2-car N, followed by K and L.”

    If a car arrives while the other is stopped: “Now boarding, 1-car J behind the K” (one possibility)

  • justin

    What about the even more logical step of [de]coupling trains at West Portal and Duboce? The fact that announcements and platform displays refer to 2-car trains as “N N” makes me think that was intended all along, like an “L M” that would decouple outbound at West Portal.

    I remember reading somewhere that the Bredas’ couplers didn’t handle this well — were they supposed to? If so, why aren’t they repaired or replaced at the manufacturer’s expense?

  • Anonymous

    I think one other reason for the ending of decoupling was the fact that it held up the line. Not sure how long the process takes.

  • Anonymous

    Okay. Let’s assume all trains are equally frequent (which is near enough the case), and in random order (which is also near enough the case). Now let’s imagine two trains are approaching together, close enough that the rear one would be delayed by the one in front if there was only one boarding area.

    The probability of your designated boarding zones reducing the stopping time for the rear train is 24% (6 out of 25). If any train could stop at any boarding zone, the probability would be 100%.

    Good electronic signage is the key to making this work; there should be a list of next trains for each boarding zone (four in total) so people can check where their train will be stopping and go to the appropriate area. The signs should update as soon as it’s known that one train will need to stop behind another, which should be as soon as the train leaves the previous station. If done properly there will be plenty of time for people to get to their train.

  • Anonymous

    So all trains have to stop twice at each platform? Sounds like a recipe for slowing things down, not speeding them up, especially during off peak hours.

  • Anonymous

    “Now approaching: 1-car J at boarding area A, followed by 2-car N at boarding area B, 1-car L at boarding area A.”

  • The quality of the comments on this site is just declining rapidly. Don’t I know what Muni’s announcement is? but it has no relevance to my concern. Currently, all trains stop on the same position on the platform. Somebody who is disabled and needs to board through the first door can always wait on the same spot and get on the train quickly. The announcement about the train length is helpful for those who is able to walk quickly and want to avoid the crowd, but it can be ignored.

    Under the double berthing situation, if there’s no designated spot for certain line, a disabled person may have to move a couple hundred feet up or down the platform at a moment’s notice, depending on when the computer thinks the train should stop where depending on the congestion level. The other alternative is to make double stop for the disabled.

  • Read jonobates comment. If the order is backwards of what is needed to match the boarding areas, the system would get *slower* – not faster. Simple math.

  • I think human factors should be included in the planning and design of double berthing. I fully support it but I think that it should be highly predictable, consistent, and disabled friendly. I value predictability and consistency over flexibility. If it is not so, some riders will end up delaying the trains anyway while making the system less safe.

  • Mario Tanev

    They don’t have to stop twice. But if there is a train in the front, the train in the back can offload its passengers instead of just idly waiting. If there is no train in the front, then there is no need to stop twice.

  • PaulCJr

    That’s the way it was design to work and at one time it did. The Breda car couplers are very sensitive due to the electronic connection also being right there, so if they are coupled too hard, the two trains will have issues communicating.

  • vcs

    Muni drivers used to do this informally, but it required switching the train out of “auto mode”.

    For inbound Embarcadero trains, they could just roll right through and not stop twice.

  • Jeffrey Baker

    This is nice and everything, but doesn’t address the problem in a practical way. I’m sorry to have to point out individual humans as the problem, but the four useless goldbricks Muni employs to move the out-of-service inbound trains off the platform at Embarcadero are causing almost all delays in the inbound subway. Those guys will stand there talking on their cell phones with their girlfriends while thousands of people wait on the platform for another useless L or M train to get the hell out of the way. Meanwhile you can see the trains backing up all the way to Powell. It’s a minute-long affair at best, and these operators have no particular urgency about it.

    They need to figure out how to have the regular operator move the inbound trains into the yard and then be relieved, not rely on these useless yard operators.

  • Jeff

    Will be interesting to see how (if) this affects travel time. Sounds like it will affect boarding times, but not travel time (granted, boarding is a large portion of transit time).

    On another note, this will be a headache disaster for tourists, or anyone not familiar with the already disorienting muni

  • Henry

    They also occasionally do this now too, but only after a couple of minutes and only the front door of the first car is opened.

  • Jamison Wieser

    Coupling and de-coupling trains at the portals creates a time hit and causes delays if the trains don’t arrive at the same time.

    Once the subway switched to a rolling block system that could run trains closer, faster and safer. Trains can enter the tunnel as soon as there is clearance, and doesn’t have to wait for another train to couple with.

    Our current fleet of Breda trains were not designed for rapid coupling and de-coupling or to run in three-car consists.

    Better train control software obviated the need to join trains. Ever growing demand makes the time-hit to couple even less affordable today.

  • Sprague

    Glad to see Muni is again attempting to improve its speed and reliability (in keeping with its “transit first” mantra). I hope the roll-out of this test/project is not subject to seemingly ridiculous delays.

  • To do that effectively you will need a staff person standing on the platform to interpret the sign and to direct people. If Muni doesn’t want to spend that kind of money then it is the best to designate a boarding zone valid at all times to minimize the confusion. Yes it is less efficient, but then we need to take account delays and complaints of missed trains caused by confusion.

    Not everybody can read and interpret signs easily. That includes older folks who aren’t so adapt to new gadgets and people who have limited English skills. There are folks with limited mobility that cannot move as quickly to a new boarding location.

    There are benefits of having two designated boarding zones. One of them would be the crowd waiting for the train would be spread further down the platform rather than concentrate in one area.

    One of the factors for successful transit systems is consistency, which means riders can easily get into a routine and don’t expect to change it constantly. Muni’s already a step behind with the difficultly of keeping up with the schedule and constantly changing boarding location won’t help either.

  • kdub2000

    Why don’t they also try to make all metro trains two-cars long???

  • Why does the inbound N stop a full car length away from the exit stairs in the Powell station, yet the outbound stops right at the exit?


Muni Double-Berthing Still Delayed Pending CPUC Approval

The ongoing delays for double-berthing in Muni Metro stations continue, as Muni waits for its training plan to be approved by the CA Public Utilities Commission. After Muni officials demonstrated double-berthing for the CPUC in December, expecting the green light, Muni Operations Director John Haley told Streetsblog that CPUC needed to sign off its training plan, […]

Muni Metro to Launch Double-Train Loading, Three-Car Trains in October

Muni is poised to begin double-train loading and running three-car trains at underground metro stations to speed up boarding and increase capacity on the system’s busiest stretch. As we reported, Muni began testing double-train loading, a.k.a. “double berthing,” in July, but Transit Director John Haley said the agency hasn’t run tests with actual trains yet. […]

SFMTA Board Approves Contract for New Fleet of Muni Metro Trains

The purchase of Muni’s next metro train fleet took a major step forward today as the SFMTA Board of Directors unanimously approved a manufacturing contract with Siemens. Muni officials lauded the design of the new trains as far superior to the current, abysmally breakdown-prone fleet of light-rail vehicles, which were built by AnsaldoBreda. The fleet […]