06 June 2009

Build Me A Train Box

The San Francisco Transbay Joint Powers Authority (TJPA), charged with building the new Transbay Transit Center, is now pursuing $400 million of federal stimulus funds to build the underground "train box" that will eventually serve as a downtown station for HSR and Caltrain.

The train box had been descoped in a phased approach adopted in June 2006, whereby the above-ground bus terminal would be built first in Phase 1, followed by a "top down" excavation of the train box below in Phase 2. This plan deferred the greatest expense until later: Phase 1 was estimated at $1.2 billion, compared to Phase 2 at $3 billion--for which few funding sources were identified. The availability of stimulus funding, and in particular an $8 billion slice allocated to high speed rail, has altered the TJPA's equation.

In its next board meeting on June 14th, the TJPA board will formally direct the design team to proceed with the train box included in Phase 1. Train box finishes (tracks, platforms, escalators, etc.) will be added later in Phase 2.

Thinking Inside the Box

The good news is that building the train box now undeniably saves money in the long run, since digging a cavernous hole underneath a fully operational bus station is no mean feat of engineering. Advantages of building the train box in Phase 1 include:
  • $100 million of construction costs saved
  • The opportunity to locate HVAC systems below grade, freeing up ground-level space and circulation
  • No difficulty with shifting foundations as the train box is built
  • Easier waterproofing (in an area with a high water table)
  • Faster construction timeline
  • $12 million of design costs and 4 months of design schedule saved by carrying only one option forward (dropping the "top down" option) allowing the project to reach "shovel readiness" before the stimulus funding deadline
  • More jobs in a time of economic recession
The bad news is that rushing the train box component of the project to achieve "shovel readiness" threatens to lock in certain detailed design features that may later turn out to be ill-advised when someone actually tries to build an efficient, functioning, workable train station inside the box in Phase 2.

The problem basically boils down to reinforced concrete columns. The entire weight of the above-ground portion of the terminal building is carried into the train box foundations through a dense forest of concrete columns, typically 5 to 6 feet in diameter, spaced every 42 feet. These columns were placed in such as way as to prevent a reasonable layout of the tracks entering the station. As discussed in Focus on SF Transbay Transit Center, the TJPA's design for the station's "throat" is a disaster that urgently needs review by competent rail professionals before any concrete is poured... and the columns peppered throughout (see diagram at left) will literally set the design in concrete.

Building the train box and its concrete columns now, while deferring the detailed design of the train station to phase 2, presents a high risk of permanently screwing up the station design. By the time the TJPA gets serious about rectifying the design of the train station, it will be too late.

A Coordinated Plan?

The TJPA's agenda item mentions a Peninsula Corridor Investment Strategy, presumably hammered out among the TJPA, the Metropolitan Transportation Commission (MTC), Caltrain and the CHSRA, to keep inter-agency conflict at a minimum in the methodical pursuit of stimulus funding. This strategy was likely included in the Bay Area Council's Economic Recovery Work Plan, which is said to request $3.4 billion for Bay Area high speed rail improvements under the following headings:

192: Caltrain electrification (~ $1.5 billion)
193: Caltrain positive train control
194: Caltrain San Bruno grade separation (told you so! ~ $300 million)
195: Caltrain North Terminal station improvements
199: City of San Jose, Diridon Station "Grand Central of the West" (~ $500 million)
205: TJPA Phase 1 train box (~ $400 million)

The dollar amounts are rough estimates, since they are not mentioned in the Recovery Work Plan. If the Bay Area secures even a fraction of this funding, get ready for quite a bit of construction, undertaken without a single cent of Proposition 1A high speed rail funds. Who would have guessed, last November?


  1. Why do the trains have to be under the bus station?

  2. Not from CA but enjoy your blog. Denver is about to begin construction on our train station and we're having many of the same problems (stimulus funds, being rushed, not enough throughput, a stub-hub). Although its outside of your region, I would love to hear your analysis on the operations side. I am afraid our station's expansion will be curtailed.




  3. Adirondacker: there's really no other place in downtown SF where you can put a train box of sufficient size conveniently, at least not that I could easily find.

  4. @ Adirondacker ... because you have to take a bus to get to Oakland.

    Obviously in a sane plan, they would find a good place to put the trains first, and then build the bus station on top of it, but the TBT decision was made on Backwards Day.

  5. Arcady, I know the satellite images are hopelessly out of date. Looks like there's blocks of parking lots right across the street from the bus terminal. 6 tracks is barely adequate, assuming the traffic projections are close to accurate.

    The Bay Bridge is at or near capacity during rush hour as is BART. Where are they going to put the trains when the new tunnel from the East Bay is built? They can't run many more trains on BART so they are probably going to build a commuter rail system that longer distance trains from places like Sacramento and Stockton can use. Where they gonna put 'em?

    Obviously in a sane plan, they would find a good place to put the trains first, and then build the bus station on top of it.

    The Southern Pacific passed on putting the train station in a good place in 1906. Doing it in 2009 means some compromises.

    Having separate bus ramps to the bridge has it's charms. Is there another place they could put bus ramps? .. they are keeping the bus ramps aren't they?

    Between Main and Beale "south" of Mission looks good to me. Connect to the bus station across Beale. If it's done right you'd barely know you are crossing a street when you change from the train to the bus or vice versa. Spreads out the pedestrian traffic coming down from Market.

    Three trains an hour during rush hour to the outer suburbs is conceivable. Three an hour to/from Santa Rosa, two an hour to Calistoga, three an hour towards Sacramento, three an hour towards Stockton and three an hour towards San Jose via the East Bay... Throw in one or two an hour from the ends of the system at Sacramento and Stockton, that's 12 trains an hour. . where they gonna put them?

    They might actually be able to run them into the six tracks under the bus terminal. I don't know if you be able to stop to let passengers on or off. And if you did if they would run into pedestrian traffic jams in the terminal.

    Doesn't really matter to me. I go to visit the relatives in Santa Rosa every year or so. When I want to go from there to see the relatives in Mountain View I can just borrow a car and drive. Since this is a pleasure trip I can do it off hours and if it takes 30 minutes longer because of traffic it's not like I risk losing my job as a cousin.

  6. ARUP is doing the Transbay Terminal Engineering. Someone there has to understand a 3 track tunnel is wasteful, and 500' radius curves are crazy, right? I'm sure with the money saved by reducing the tunnel to 2 tracks they could do the necessary extra construction and engineering to reduce the curve radii some.

  7. What is the design speed of a 500' radii track alignment?

    750' radii is 25mph... from my understanding. 500' might be 15mph???

  8. Nobody with even the most rudimentary familiarity with rail operations -- even at the level of running a model train set -- was ever or is now involved in the planning of the Transbay Terminal rail level or of the Downtown Extension.

    The unprofessionalism and the ignorance is breathtaking.

    Caltrain, the only agency that could have had influence over the retarded children from Parsons Transportation Group (the TJPA's consultants -- and Caltrain's Electrification Experts for that matter) appears, as far as anybody outside can tell, to have washed its hands of the downtown extension years ago.

    The Transbay Terminal building itself was "designed" by some civil engineer who took out his trusty inches-and-feet CAD system, decided that a 36x3 rectangular grid of columns spaced 42'6"x52' fitted within the footprint of the real estate, and THAT WAS IT.

    The lateral column spacing doesn't even line up with the centrelines of the platforms, let alone allow free movement to and from trains. The longitudinal spacing doesn't allow space for even a single stairway or escalator to have a clear run -- instead narrow, single escalators (which are UNIDIRECTIONAL, note!) are placed next to five foot diameter, 42'6" spaced, immovable, inhuman, claustrophobic building posts, blocking circulation and even further degrading efficient train operation.

    Nobody at any stage in any way made any effort to even verify that this rectangular grid of columns was in any way related to or compatible with or detrimental to the running of trains into what is nominally sold as a FOUR BILLION DOLLAR railway station project.

    The "architectural" (!) "design" (!!) "competition" (!!!!!!!) was a complete sham matter of slapping some lipstick on that pig. Literally! Everything in the "winning" "design" is facade over an unworkable, obstructive, brain-damaged, unchanged, unchangeable array of concrete columns. The structural grid pulled out of somebody's ass seven years ago has been changed only in the very slightest, most trivial way. (Make way for the skylight oculus! Don't even think about human movement.)

    The result -- including a dank mezzanine with holding pens, including passenger "circulation" in which every single pedestrian route to and from the trains run right into a column, in which there is simply no architectural clue that a train station is even present under the lipstick-smeared above ground pig of bus station, in which there is no room for stairs to and from the platforms because columns block the way, makes it clear that New York's abysmal Penn Station is the only model with which the profoundly ignorant, profoundly incompetent, and profoundly unprofessional "engineers" and "architects" involved have any familiarity.

    A good deal of attention seems to have been paid to ensuring that the bus terminal works (which it well might), largely because somebody from AC Transit made an effort to look after the interests of bus passengers, but there has been less than zero rail planning, at least by anybody who knows anything about running trains or moving passengers.

    The result is an absolute, umitigated -- and totally avoidable -- disaster.

    This is far worse than BART to Millbrae (which only put back transportation in the Bay Area by 30 years) or HSR via Pacheco (which will only put back rail transportation in Northern California for 40 years) -- the unlimited unprofessionalism of everybody involved with the TJPA will cripple regional rail service into San Francisco for the next century.

    Those responsible -- directly, or through negligence -- should be put up against a wall. There is no excuse.

  9. Re: "500' might be 15mph???":

    Bad as the speed is, that's not the worst of it.

    A 500' (164m) minimum radius is below the standard revenue service design minimum of every first world mainline rail operator.

    Think what that means for vehicle design and procurement, just for starters, even ignoring the disastrous cumulative effects upon operating efficiency.

    And it gets even worse. Here's some actual "logic" from actual TJPA "professional" consultants who are doing the actual "engineering" on the putative railway station: "Although CHSR specifies that all platforms should be located on tangents, there is no minimum radius specifically for use in platforms in the DTX criteria, so the minimum radius of 500' should be applied."

    As I said, the unprofessionalism and the ignorance is breathtaking.

  10. So is this money going to be on top of the request that the CAHSRA submitted to the FRA? If so, how sweet would that be.

  11. The CHRSA doesn't get to submit anything to the FRA. Everything goes to Will Kempton and Caltrans to be the final arbiter of the list for FRA.

    Should get interesting since the CHSRA list includes the whole SF-SJ segment but the Bay Area working group (a group that contains CHRSA) does not.

  12. I have seen your writings before, and I am glad you point these things out.

    Platforms on 500' radii curves... ASTONISHINGLY SHORTSIGHTED! How far from the platform will the doors of cars line-up? 10-feet or so? This reminds me of high school trig. ...and, did a high school student do this work?

    Eric... the CHSRA has not submitted any application for stimulus money yet. And, don't fault them. Fault teh TJPA for the under-design of the TBT... and for having blinders on concernign the deficiency of the design. The CHSRA is a victim in this (the public too).

  13. Brandon,

    I am not faulting anyone. I realize that the requirments for getting a share of the 8+ billion dollars for "high speed rail" have not been established yet and are due to be released sometime this month. I just noticed the list from the Bay Area Council Economic Institute are different (dont fully include) what CAHRA wants. I see on Roberts Blog, the same topic is starting up too with his Sunday open thread.

    By the way, I enjoy the technical discussions on this blog, Clem Richard, others, as well as the not so technical on Roberts blog. Both well done!! I read a lot, but don't post much unless I get a bug up my butt, or there is a comment from someone in right field that I feel needs to be addressed

  14. This building will be a transportation nucleus for years to come and it needs to be built right. How can we communicate these deficiencies before the structure gets built?

  15. @ FAS -

    if anon @ 19:28 is correct, the appropriate mechanism would be for Caltrans to hire a small team of consulting railway engineers with members from the US, Europe and Japan.

    Their initial task would be to double-check the plans that currently exist for phase 1 to verify that they will permit a phase 2 railway station that meets train operator, pedestrian flow and other requirements.

    If Richard Mlynarik is right, and I believe he is, that design review will conclude that there are major problems.

    A larger number of substantially more slender columns based on steel girders encased in a relatively thin layer of concrete might be one way to create space for sufficiently wide, straight sloped walkways (cp. e.g. train station under Amsterdam's Schiphol Airport).

    Various technologies, e.g. steel space frames or beams of pre-stressed bridge-grade concrete could permit wide spans above the station throat. Even better might be an above-ground arch that transmits the weight of the upper floors to supports on the sides of the throat. The street-level area could be a pedestrian plaza suspended from that same arch. The downside would be a loss of some indoor square footage.

    When you build a train station, the train and pedestrian flows need to be the primary considerations. Shops, food courts etc. can be moved around much more easily.

  16. Maybe this "design" will be in place until they get the money: "Hey this thing is shovel ready!".

    Then "last minute" changes could be made.. "opps hey look at these problems that we found. Who knew that 500' radius was too tight? Looks like construction will have to be delayed! -wink wink".

  17. "Although CHSR specifies that all platforms should be located on tangents, there is no minimum radius specifically for use in platforms in the DTX criteria, so the minimum radius of 500' should be applied."

    That's... ridiculous. Some quick trigonometry shows that with 85 foot cars, there will be a 1 foot gap between the car body and the platform at the center of the car on the outside of the curve. Of course, nobody really knows what height the platforms will be, and what design criteria should be applied to ensure an acceptable platform-train interface.

  18. I think the column locations would be set even with the old phasing. Either way they'd better figure it out first. It might be easier to put in thinner columns, or transfer beams if it is done all at once.

  19. This comment has been removed by the author.

  20. "As I said, the unprofessionalism and the ignorance is breathtaking."

    Richard, do you have a mental condition that causes you to be disappointed with everything everybody does? I mean, you have a problem with Muni Metro, BART, VTA, the CHSRA, and even Smugmug.

    This is not meant to flame you because I am genuinely curious. Does anyone or any organization do anything right in your opinion?

  21. Spokker, Richard has expressed admiration for the way trains are run in non-English speaking countries, singling out Germany and Switzerland for praise.

  22. Funny, that, because when Switzerland and Germany use 16.7 Hz power supplies, that's somehow okay to Richard, but when Amtrak uses 25 Hz, that's "antiquated" and "stupid".

  23. LOL (at Spokker's link), Richard even hates the fact that a site might require Javascript. That's pretty hardcore...

  24. Amtrak doesn't use 25 Hz; the Pennsylvania and New Haven Railroads did. The new catenary north of New Haven is 25 kV AC, 60 Hz.

  25. Alon: South of GATE interlocking in Queens, the current in the overhead wire is AC at 25 Hz, which is what powers the Amtrak trains from there south to Washington. Maybe that doesn't count as "use" to you, but it certainly does to me. It's true that north of there to New Rochelle, it's 60Hz, and north of New Haven it's 25 kV as well, but Amtrak has also been upgrading their frequency converters in the Philadelphia area, which is a pretty good sign that they intend to keep the 25Hz system around for a long while yet.

  26. The salient point is not how many things Richard is critical about, but whether the critique of the TBT train box has validity.

    It appears that it does ... even a non-engineer can trace through the rail movement bottlenecks in the TBT design, which Richard's design eliminates, grasp the fact that there is no extra capacity gained from the third track through the tunnel, and understand that larger platform curve radii are superior to tighter platform curve radii.

  27. Bruce: I completely agree. If even Richard Mlynarik can design a better track layout for the TBT, then there's something very wrong with the design.

  28. @ bruce " but the TBT decision was made on Backwards Day." that's funny. but seriously, the tbt was planned for san francisco to replace the old tbt- the train box was added for the caltrain extension and "possible future high speed rail" to the rest of the world it always looks like "backwards day" here in the city, but you know. love it or leave it. try it, itll grow on you.

  29. as for all that lipstick - it was the selling point for the people of sf, we would not have allowed a project this large and this expensive and this potentially disruptive to residents without all those perks in the design. I doubt anyone around here cares if there are columns on the platforms as long as the park is nice and the trains run on time and you can get organic sushi and microbrews there. Not saying thats the best approach, just telling it like it is.

  30. jim said...
    "... seriously, the tbt was planned for san francisco to replace the old tbt- the train box was added for the caltrain extension and "possible future high speed rail""

    But the "possible future HSR" was not added in the design phase, its a mandate from the '99 proposition. And while the people who designed the TBT tunnel and train box might not have taken that mandate seriously originally, because they thought that the HSR would never get funded before the TBT got up and running ... those assumptions turned out to be false.

    So they have a responsibility to go back and fix it up so that it is genuinely designed to serve as SF's HSR terminus, as the voters of SF dictated when they approved the proposition.

    500' turn radii and a three track tunnel with no better throughput than a two track tunnel and with worse cross platform interference doesn't cut it.

  31. porp H does say high speed rail but doesn't go into detail as to how many trains or what the box should look like tis does specifically say "(e) result in the lowest feasible combined costs for construction of the bus terminal and the Caltrain station, without sacrificing the aesthetic qualities of the terminal and station and their interface with surrounding development."

    the point being, do what must be done but not at the expense of the the rest of the project.