The Current Situation

The Transbay Tube in the San Francisco Bay Area is the longest underwater tube in the world and an impressive engineering feat. When it was completed in 1970 it was also expensive: the Tube had a price tag of $180 million. Unfortunately the tube is not meeting the current capacity needs of BART, the train that uses the tube. Building a new tube has been discussed, but this new tube would be more expensive than the original, costing over $10 billion by some estimates. For California, a for state that fraught with budget problems, this is a prohibitive sum of money. To make matters worse the Eastern span replacement of the San Francisco–Oakland Bay Bridge has been plagued by cost overruns and political infighting, which has spoiled the political environment for large infrastructure projects such as a new transbay tube.

A Possible Solution

Fortunately, a new tube is not the only way to meet BART’s transbay capacity needs. Currently train capacity through the Tube is limited by headway time: the amount of time trains must spend at stations; a new train can’t come into a station until the one in front of it leaves. Technically BART can run at two minute headways, but loading and unloading passengers takes additional time, so longer headways are needed. Adding more doors to future train sets may help this situation, but this alone won’t be enough to address future capacity requirements.

However, if the stations could handle twice as many trains – if there were four tracks through San Francisco instead of two – technically the existing Transbay tube could accommodate twice as many trains as it currently does. Under Market Street, where the Transbay Tube comes into San Francisco there is not sufficient space for four tracks, but one block south on Mission Street a new line could be built. This station would have the added advantage of being part of the new Transbay Terminal (TBT).

Additional Benefits
There are other benefits to this proposal. First of all this project can be completed in phases. That means, unlike a new transbay tube, relatively small investments – but investments with immediate results – can be made over a period of time as funding becomes available. This project would also work in coordination with the new TBT: a two line solution in San Francisco would allow a BART station to be built in the nearby Ferry Building thereby linking not only BART, but the TBT to this facility. The new station at the Ferry Building would also reduce the number of passengers using the Embarcadero Station. This is important because the Embarcadero BART station is the busiest BART station and the cause of longer headways and delays. By placating this bottleneck this proposal could more than double BART’s capacity. Connecting BART to the TBT would also link the TBT, the planned terminus for HSR, to the SFO airport via the BART-SFO extension. In subsequent phases a second BART line in San Francisco could link important destinations such as Moscone Center, Herba Buena Gardens, the Central Subway, the Van Ness Muni Metro Station, and City Hall. This line would also help ease congestion on the Van Ness corridor, where a BRT study has already been undertaken and in the past tunnels have been proposed to re-route traffic.

Technical Solutions
Stay Tuned… we’ll have more information soon that addresses the technical hurdles associated with this proposal.

Photo Credits
Image 1: BART by Lee Comma Dennis (Flickr)
Image 2: transit map by SFMTA
Image 3 & 4: Transbay Tower and Terminal by Pelli Clarke Pelli Architects
Image 5 & 6: Google Earth Images by Brian Tyler

© Brian A. Tyler and, 2009.
Unauthorized use and/or duplication of this material without express and written permission from this website’s author and/or owner is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Brian Tyler and with appropriate and specific direction to the original content.

17 Responses to “BART: increasing system capactity without a new transbay tube”

  1. 1 anonymouse April 21, 2009 at 6:59 pm

    I don’t think BART has really used the full potential of their existing subway and tunnel infrastructure. The first problem, as you mentioned, is dwell time at Embarcadero, and getting trains with more doors would go a long way toward fixing that. There might need to be some vertical circulation upgrades, but I don’t think overcrowding is too much of a problem just yet. Also, they don’t even run all 10 car trains during rush hours: changing that would boost capacity a bit. Building a tail track at Civic Center or so would help turnback capacity too, and all together you could probably get 32 10-car trains per hour, if not a bit more, depending on how reliable merging is at the Oakland Wye. Now the challenge becomes how to pack as many people as possible into each train car, and the way to do that is to replace seating room with standing room, possibly by changing transverse seats to longitudinal, and probably by making bigger standing areas. It might not be unreasonable to fit up to 250 people per car at crush load, for a total transbay capacity of 85,000 compared to current actual capacity of around 33,500.

    Your proposal might be a way to increase tunnel capacity to 60 trains per hour, using some very optimistic assumptions, including no stations at all on the common segment (that includes West Oakland). Of course then there’s the problem of the feasibility of building it all, especially the junction to the existing line, possibly under water.

    • 2 switchingmodes April 21, 2009 at 8:16 pm

      Good points, all of which will be addressed in the “Technical Solutions” part of this page. However, I’ll give you a brief preview: 1.) Four tracks through Oakland including grade separated cross-over tracks and a two platform, four track station at West Oakland Station where trains would be staged at one minute intervals. 2.) The objective of one minute headways would probably require an all new cab signaling system. This would be part of the fleet replacement that BART needs to undergo anyways (which six doors would be a part of, and necessary, to achieve these headways) 3.) BART would probably have to shut down the TB tube for at least a three-day weekend (like the new East-span Bay Bridge replacement occasionally requires). A new Jack London Square multi-modal station, and a redirected fleet of Ferries would be used to shuttle passengers across the bay during these periods. Also, an ‘experimental’ bus lane on the Bay Bridge would be nice, if it could be pulled off 4.) The inevitable periods of TB tube downtime would be reduced by using prefabricated sections that would be sunk into the Bay to replace some of the existing prefabricated sections used in the original tube. 5.) There would need to be a new four track, two platform station at the Ferry Terminal to allow transfers between lines and trains to be staged at one minute intervals. The first platform constructed would be for the new BART line. Once completed, the Market Street line could be temporarily shut down and the new line would be the only BART line through the city until this could be completed. This new Ferry terminal station could also prefabricated and sunk into place. If funding was not available to allow the new line to head all the way down Mission/Howard to connect with the old line when it head south of Market, the TBT would temporarily serve as a terminus station (under Mission St.) and a walkway would be used to connect with the Market Street BART line.

      I do have some issue with you comment though: 1.) Removing seats for more passengers: some BART rides are very long and this could deter many riders and I don’t think the figures you presented are realistic. 2.) 32 trains per hour and more people on each car: What ultimately holds BART back is headway time which is increased by passenger loading and unloading at busy station – putting more people on each train would decrease the number of trains that BART could send through the tube 3.) Turn back capacity at Civic Center: this would not increase capacity unless you are considering that BART could then run longer trains because more train cars would be available. In that case, this is an expensive solution. I think just buying more train sets would be cheaper 4.) Overcrowding is not yet a problem: I think overcrowding is a problem and is deterring transit riders. Also, this new TB proposal takes a long-term approach that would be completed in phases. While I think it could be completed faster than building a new TB tube, we’re looking at over a decade if we start today and meanwhile the congestion issue is not getting any better.

      • 3 anonymouse April 21, 2009 at 11:37 pm

        First off, the cab signal system is not actually as big a deal as signal vendors would like you to think. There are real fundamental constraints in terms of braking distance, how fast switches work, dwell time at stations, and schedule reliability especially at merge points. Moscow uses an even simpler system than BART, and gets 39 tph (reliably) on its busiest lines, and interestingly enough they don’t use ATO because it would lower capacity. And by the way, Moscow has about six lines that carry more than 1 million daily unlinked trips _each_. As for seats, I think comfortable seats for the long journeys of BART are certainly desirable, but I think being able to get on the train is more important than having a comfortable seat during rush hour, and outside rush hours trains aren’t exactly full most of the time. Oh and my comment about overcrowding was only meant to refer to the station platform at Embarcadero, not overall, sorry if that wasn’t clear. I do agree that to some extent ridership is being deterred due to crowding and something needs to be done about that. The only purpose of the turnback at Civic Center is to increase reliability by reducing the length of the section with service being run to its limits to Oakland Wye-Civic Center rather than all the way to the north Daly City crossovers. It also would be much, much better than the current arrangement with turnbacks at Montgomery.

        From a technical standpoint, I think the key difficulty in your proposal is connecting at the San Francisco end. I think you’d need to build a flying junction underwater, probably using a cofferdam to provide a work area, which sounds like a tremendously expensive proposition, but quite possibly not as expensive as a new tube. If you can do it, I’d really love to see a proposed alignment for the junction, as that would really give the whole idea some plausibility.

  2. 4 Erik April 21, 2009 at 11:31 pm

    Is there a reason that the lower deck of the Bay Bridge could not be used to increase capacity across the bay as well?

    • 5 anonymouse April 21, 2009 at 11:46 pm

      Can’t take lanes away from cars. Otherwise, it might not be a bad idea, perhaps to return something like the original Key System which used the tracks on the bridge. They ran two-unit trains with articulated units that looked quite like the modern LRV layout, and had very close headways. Unfortunately, their top speed was only 35 mph and there’s a very good chance that there will be a speed restriction on the bridge of 40 or 50 mph, rather than the 70 that BART does in the tunnel.

      • 6 Erik April 22, 2009 at 11:57 am

        I just don’t understand that mentality I guess. I mean, reducing service on a transit system is an open possibility when agencies are fiscally strapped, but taking lanes away from cars is absolutely not ever allowed for some reason. The fact that converting a few lanes of that bridge to public transportation services would increase the capacity of that bridge immensely. And it seems like converting the bridge and upgrading it to allow for higher BART speeds would still be far cheaper than building a completely new tunnel under the bay.

    • 7 Brian S April 22, 2009 at 2:26 pm

      Yes you could use the Bay Bridge for added transit, however not for rail. Instead you can implement a contraflow lane on the bridge and on freeway approaches ton vastly increase transit access and capacity in the Transbay Corridor.

      By placing one or two contraflow lanes on the lower deck in the AM peak, and on the upper deck in the PM Peak, vehicle capacity would remain roughly the same in the peak direction, with added transit capacity, while only mildly affecting off-peak direction traffic.

  3. 8 Brian S April 22, 2009 at 2:32 pm

    At the Lincoln Tunnel in New Jersey, one lane of the AM westbound (off-peak) lanes is given to contraflow buses heading eastbound into Manhattan. This has worked for decades. Consequently, the buses passing through the Lincoln Tunnel carry more people from NJ into NYC than PATH and NJ Transit rail combined.

    We can do it on the Bay Bridge. If you show that not throughput of traffic is affected for AM eastbound and PM westbound, there’s no reason not to do it.

  4. 9 JoeC April 22, 2009 at 9:51 pm

    If the bottleneck is off-loading rush hour passengers, is it possible to run a skip-stop service for the four Market Street stations and double up the frequencies? So every other train would stop only at Montgomery and Civic Center, while the next one (at one minute headways) would stop only at Embarcadero and Powell. Basically the two trains would be offloading twice as amny people at each station, but at only half the stations. I guess at some point some of the trains would have to turn back – which might create a bottle neck further out.

    • 10 switchingmodes April 23, 2009 at 8:40 am

      Your idea would work (sort of), if trains could pass. The problem is the demand is at Embarcadero and Montgomery. So, even if trains could pass to avoid those stations you would have an even worse situation then there currently is for those trains that do stop there. This would also create another bottleneck: passengers would then take BART back the other direction to the station they need. Furthermore, such a plan would be confusing for passengers and logistically very challenging. It would also not be especially effective because currently BART is limited to two minute headway times in the best of situations.

      The approach presented here tackles the problem from many angles. It spreads out the Embarcadero Station station demand out by building a new station at the Ferry Terminal and by building new stations South of Market. It then allows doubles the trains per hour by running trains very close together – think of it like having 20 car BART trains – which can be done because those trains head to different stations. If there where six door trains and excellent logistical and technical planning, I think the TB tube could have one minute headway times. If you add to that some less seating in new articulated train sets, we’d have something just as good, if not better than, a $10b + new TB tube.

  5. 11 anonymouse April 23, 2009 at 1:22 pm

    Come to think of it, if Embracadero is the bottleneck, have they really optimized the signal blocks on the approach? Could they reduce dwell times by having station agents on the platform moving people along? Paying 10 people’s salaries is MUCH cheaper than building a new tunnel. And I still think BART can cut the number of seats, or maybe just make them smaller. They really are luxurious even by East Coast commuter rail standards, like having a couch. Oh, and rip out the carpet from trains. It’s responsible for way too much maintenance downtime.

  6. 13 Rafael September 21, 2009 at 7:24 am

    “Technically BART can run at two minute headways, but loading and unloading passengers takes additional time, so longer headways are needed.”

    Sounds to me like the focus should be on improving pedestrian flow capacity inside the trains, on the platforms and up to the surface. Switching to rolling stock with additional doors and fold-up seats along the sides would be a start. Adding side platforms to the existing Embarcadero and Montgomery stations would involve cutting holes into the existing tunnel walls exactly where the doors of the trains are. For obvious reasons, trains would have to stop at the right location. The side platforms would connect directly to new surface exits west and east of the existing ones, bypassing both SF Muni subway and the shared concourse level. All BART trains would open both sets of doors at these modified stations.

    Another, possibly more difficult, approach would be to persuade SF businesses to stagger their working hours or, to relocate some back office operations to the East Bay.

    Spending $10 billion or more on a new transbay tube and BART line down Mission Street may become necessary someday, if only for seismic redundancy, but not yet.

    Bitter irony: Quentin Kopp nixed plans to include support for future BART/light rail/HSR tracks on the new east span of the Bay Bridge because a project to add cantilevered tracks to either side of the lower deck of the west span plus tunnels through Yerba Buena island plus approaches into SF would have cost $3 billion.

    Now, SF will have to hope that the extra-fancy new TTC will provide the requisite additional capacity via buses alone. Articulated bi-level motorcoaches might help, but only if there are enough bus lanes in the East Bay to avoid getting these vehicles stuck in traffic before they ever get onto the Bay Bridge. Building ridership will depend on offering travel times that are competitive with BART (hard) and/or greater comfort (easy) and/or lower fares (depends). In addition, it could be valuable to offer free broadband internet access on board the buses.

  7. 14 Aaron December 31, 2009 at 1:09 pm

    I always figured the logical thing would be to build a new Muni Metro tunnel under Mission between 11th and Embarcadero, and allow BART to use both levels of the existing Market Street subway. A Muni Metro tunnel would have much shorter platforms and thus be cheaper to construct, and it would be easier to have just one set of paid areas in the BART system.

    Another, possibly less intensive solution would be to dig out the outsides of the platforms at Embarcadero and Montgomery. It would allow both sets of doors to open, which should speed things up somewhat.

  8. 16 Michael F. Sarabia January 21, 2010 at 8:30 am

    Why not expand the Dumbarton Bridge with a BART rail?
    Since BART will reach the Transbay limit, sooner rather than later,
    it is best not to put all eggs… etc.
    Someday, a long, long time from now, BART will find its way to San Jose, after it builds (waste $500 Million each) the Oakland Airport Connector and the Bay Point to Antioch eBART extension with a totally new kind of train with its own crews, maintenance people and parts, etc. Is anybody in charge of BART planning? Are they getting “a cut” from their bank loans?
    Why, put the system in debt for $1 BILLION dollars? Will they have to pay it back? With interest? By increasing fares? Which decrease the number of minimum wage workers than can afford BART?
    How many minimum wage workers from Pittsburg/Bay Point could afford to continue working at SFO with a daily roundtrip ticket at $21.80?
    Do they care? They got their pension by now…

  1. 1 BART: increasing system capacity without a new transbay tube « Trackback on April 23, 2009 at 7:06 am

Leave a Reply

© Brian A. Tyler and, 2009.
Unauthorized use and/or duplication of this material and/or concepts without express and written permission from this websites’s author and/or owner is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Brian Tyler and with appropriate and specific direction to the original content.

Page copy protected against web site content infringement by Copyscape


May 2010
« Sep