|Transbay: courtesy of Pelli Clarke Pelli.|
During the discussion at the Metropolitan Transportation Commission over how to allocate Bay Area transportation stimulus funds, MTC proposed applying for $195-$400 million of funds to build the train box at the Transbay Transit Center in San Francisco. This money would come from the $8 billion of high-speed rail grant money that was ultimately integrated into the final stimulus bill. Then, the California High-Speed Rail Authority made public its concerns that the Transbay Transit Center, as currently designed, would be inadequate to satisfy its new needs for 12 trains per hour, increased from six trains, for each of the six peak hours everyday. Under the current plan for the Transbay station, tracks would transition underground west of 4th Street, leading to a new subway station at 4th and Townsend, where Caltrain currently terminates. A three track tunnel would then curve off Townsend and north onto 2nd Street, turning once more and splaying out to a six track throat leading to the Transbay underground rail station. The station had been planned to include three island platforms and six platform tracks: two tracks for Caltrain, and four for high-speed rail. According to ridership projections, 2030 Transbay ridership for Caltrain might be 31,500 and 4th/King Caltrain ridership at 17,100. Daily high-speed rail ridership at Transbay by 2030 was projected to be in the vicinity of 26,500.
The basic problem is a perceived discrepancy between the CHSRA’s ridership projections and an as-yet unrevealed operational plan that would require 12 trains per hour and 8-10 platform tracks, rather than four. The Transbay Joint Powers Authority will need a detailed explanation from CHSRA justifying its demand for extra capacity. But what would it take to build more tracks? To avoid the cost of additional right of way acquisition, the TJPA considered a design which would accommodate six additional tracks for high-speed rail on a third level underneath the currently planned platform level. (The above image depicts the concourse mezzanine and the single planned platform level.) The extra cost of such a change would be at least $1 billion: $500 million for the expanded train box and $500 million extra for the actual rail extension — and that even lowballs it, because it assumes early construction of the expanded train box. Is there a demonstrated need for the extra capacity, that would justify somehow scraping together even more funding for Transbay? In Tokyo, 13 trains per hour at peak time for the Tokyo-Osaka Shinkansen line use just six platform tracks. But that’s for a line that carries 145 million passengers, which exceeds CHSRA’s projections even at full-system buildout — and that’s in a country whose current population is at least double California’s projected population by the year 2050. The TJPA also considered a worst-case scenario with a higher Transbay ridership (12.7 million annual riders), in which an especially large number of passengers happened to concentrate on peak hour trips. But even in that worst-case scenario, combining the CHSRA’s ridership projections with its new alleged need for increased peak capacity, the trains would only be 12-43% full at the newly increased service levels. Note that there are still other design options that were once considered — including tail tracks and an underground track loop — which would improve the flow of trains in and out of the station. And in the long, long, term, beyond 2030? An additional tube under the Bay has long been discussed in hushed tones, even though it is not yet being planned outright. If an additional tube contained four tracks (with two broad gauge tracks for BART, and two standard gauge tracks for intercity and high-speed rail), it would dramatically increase pinched transbay capacity. It would open up new possibilities for regional connectivity, and it would finally integrate Oakland and the East Bay directly into the state’s high-speed rail network. It would also mean that trains would run through Transbay, rather than terminating there, which would put less stress on that station.
It has been known for about eight years that the TJPA/Caltrain “design” for the TTT — basically something sketched on the back of an envelope by Caltrain’s freight design Engineering division (which is the only one they have) and the TTT EIR consultant — is grossly and laughably inadequate for any type of rail operations.
Some of the more glaringly egregious problems include:
* Insane and unnecessarily tight curvatures (below 150m) which are not dictated by the (very constrained) site, but which result in lowest possible speeds, lowest possible station throughput, and a state wide constraint on vehicle design.
* The very least flexible possible station throat design, in which adjacent pairs of tracks neck down to single crossovers — actively preventing the conflict-free parallel moves (ie simultaneous arrivals and departures) which are the most important determiner of station throughput.
* By using US freight railroad design standards (the same 1950s US Olde Tyme Railroading that make Caltrain and Amtrak the shining beacons that they are), the length of the critical station throats (interlockings in US parlance) is maximized, meaning that not only will trains be conflicting with each other all the time because of scandalously incompetent throat design, and not only will they be travelling as slowly as possible, but they’ll have to travel as far as possible in order to get out of each other’s way.
If you were to attempt to minimize the number of trains per hour that could be accommodated within the footprint of the Transbay Terminal site, the wonderful engineering of Parsons Transportation Group and the TJPA is about as close to perfect as your could come.
* A completely insane design for the Mission Bay (Fourth and Townsend) station, where they build a two-outside-platform, one centre-bypass-track structure underneath Townsend Street and retain an entire, completely grandiose ground-level Caltrain station to serve the hundreds of trains that their incompetent TTT won’t. The outside platforms guarantee minimum operating flexibility since all outbound stopping trains will have to use one platform (think about it), there is no possibility for by-directional bypassing, and the subterranean location means a miserable rider exxperience, inconvenient access, and a dismal US-typical “mezzanine level” pushing the station even deeper.
Oh yes; this station is designed with “Caltrain height” and half-length platforms — HSR trains couldn’t stop there even if they wanted to. (Which they reasonably might in the case of service disruption.)
In a professionally competent world, the entire stretch Bayshore-Mission Bay-TTT would be designed and operated as an integrated train handling facility, with Mission Bay with a particular role in ensuring that trains enter and leave the final, most congested and most critical TTT approach stretch at optimal times and on optimal tracks to maximize throughput. TJPA’s crazy little underground hole and huge, unecessary surface parking lot does neither.
* The very concept of platforms “dedicated” to HSR or Caltrain — complete with proposed DIFFERENT PLATFORM HEIGHTS and DIFFERENT PLATFORM LENGTHS. Nobody else in the world would even dream of proposing such a laughably insane concept at a site in which every inch and every second of flexibility counts.
Dedicated platforms mean that any single failure (a train that runs even slightly late, a train that fails, a passenger-induced problem) has the potential — and, at a site as physically constrained as TTT, almost the certainty — of causing knock-on delays. If a train can’t be flexibly re-routed to use whatever space is available at any given time, the absility of the station to reliably handle any level of traffic is hugely diminished.
The whole proposal is just rankly amateur and completely incompetent.
HSR and commuter trains serve the same station platforms except at stations with the most lavish amount of platform and track real estate (which TTT certainly does not have an will not have under any scenario.)
* The idea that HSR, which is the tail wagging the dog, even at a REALISTIC 4 trains per hour (let’s ignore the outside outlandishly fraudulent 12tph which the same PBQD that came up with BART Millbrae 30,000 riders/day dreamed up) should monopolize 2/3 of the station and then PARK ITS TRAINS OUT OF REVENUE SERVICE for upwards or half an hour at a time just shows a total ignorance and/or contempt for both 20th (let alone 21st) century rail operating practice and for any idea of cost control. Pork from the heavens will pay for the most over-the-top gold plating after all!
Consider than 12tph of 400m long high speed train is the equivalent of over 90 completely full 737 aircraft per hour. Where the hell could the riders to fill this sort of capacity possibly come from — other than the bulging, well-used folder labelled “systematic and deliberate contractor ridership esimate fraud”, the one to which the involved parties have had so very much use over the last couple decades of BART-extension-related fraud.
* The TJPA underground “plan” DOESN’T EVEN MANAGE TO LINE UP BUILDING STRUCTURAL COLUMNS WITH THE PLATFORMS. Apparently geniuses responsible for the “architecture” of this disaster decided that the building above would have columns 52′ (feet! FEET! Good God!) apart, and, well, whatever is underground would have to deal with that. (Note that with a reasonable 10.0m platform width, 1.5m platform to track centrleine spacing and 4.5m inter-track spacing the distance between centres of platforms is 15.25m or about 50 feet) So one ends up with a mess on the platforms in which columns lie close the the edge of platforms and in which escalators, elevators and stairs are offset and push up against platform edges. Oh, and to make the environment even more gorgeous, the lateral separation between the columns is only 42’6″ (~12.9m) which means that the entire “environment” will be one of a forest-of-columns, and in which it is impossible to even locate a single stairway or escalator on the centreline of the platform (where the must be, to provide maximum passenger circulation space) between adjacent column rows. Oh, and the insane, minimum-radius curves on which they site the platforms aren’t allowed for by the God-Ordained Structural Grid — the columns just march right up to the edge of the platform.
Obviously nobody in any way connected with the project has ever had anything to do with or even visited a contemporary, well-designed passenger facility of any type. (And I’m not talking about New York Penn Station, Philadelphia Suburban or anything else in North America.)
Besides making an unpleasant environment, inhibiting passenger circulation and lengthening train stop times, professional negligent design such as this is a safety disaster and a lawsuit waiting to happen.
But it’s what we get there in the USA with the best allowable US Engineering Consultants.
* 12 HS trains per hour is simply fraudulent. Where are the 12tph Madrid-Barcelona? Where are the 12tph Frankfurt-Köln? Where are the 12tph Paris-Lyon? Where are the 12tph Tokyo-Kyoto for God’s sake?
The people involved in developing these “estimates” and “requirements” at CHSRA have a long and uniform history of grotesque overstatement of ridership, under-“estimate” of costs, and of gold-plating projects to the sole benefit of the engineering/construction mafia.
The numbers simply don’t pass any sort of laugh test.
* MOST RIDERS ON THE CORRIDOR WILL BE CALTRAIN RIDERS. Yet Caltrain service is effectively completely shut out of the Transbay “planning”, and Caltrain passengers are to be forced to deal with the disaster that is Muni in order to reach the CBD.
The CHSRA is a tail wagging the dog. A pork-filled, rent-seeking tail.
Caltrain passengers are being screwed by a scam that leaves the most important market segment and most environmentally beneficial service out in the cold. We’re going to spend $4 billion and have most passengers left stranded at Fourth and King? Who is responsible for even dreaming about this, and how soon can they be terminated?
* The far-less-than-world-class US “transportation planning engineers” associated with the project are obsessed by “tail tracks”. These are underground, CBD-located train parking lots beyond the platforms — which would be about the most spectacularly wastefully expensive place to park trains, even if they didn’t completely screw up the layout of tracks within the station itself, by requiring another set of (crazily, constrictingly., screechingly tight) curves at the eastern end of the station. TJPA’s design has gotten so bad that they in fact cut off half of the “Caltrain” platforms in order to have the tracks make a reverse curve into the platform space so that they can reach the all-important tail tracks without running into the building at the corner of Natoma and Beale.
In the real world, trains are kept in revenue service and do not skulk around in underground sidings all day — especially not at a cost of hundreds of millions of dollars, and especially not at the cost of train and passenger space inside the station. (Besides which, given 400m-long double-HSR-compatible platforms, Caltrain could if necessary park a couple trains at the far end of platforms off-peak and still have more than enough space on the remaining half platform for normal operations.)
* The TJPA plan includes a full-plate, full-length, full-width underground “mezzanine” level, which means that the train level below will be oppressive, largely unlit, utterly absent of any feeling of “arrival”, and generally a confusing warren. Apparently the HSR feel that a huge airline style “passenger security” holding pen — complete with insane “security theatre” passenger cavilty searches and X-ray and baording pass nonsense — is necessary. And somebody at the TJPA feels that a separate holding pen for Caltrain passengers (a waiting room! what is this, 1950?) hidden behind a bank of FARE GATES (what is this? 1970?) is also necessary.
Meanwhile, well-designed, modern, passenger-oriented facilities (the newest exemplar being Berlin Haupthahnhof) feature none of this nonsense.
Airline “security” for HSR simply means infinite delays in boarding and infinite train capacity problems in the terminal — better not let SCARY CALTRAIN serve the same “secure” platforms as HSR! And of course it provides NO SECURITY WHATSOEVER, as the example of the Madrid bombings showed. (Spain has security theatre for its high speed AVE trains, but all that means is that commuter trains get bombed, with many times more casualty and death.)
Fare gates and “waiting areas” simply impede passenger movement and make the whole point of a downtown rail extension, which was once supposedly all about saving trip time for passengers and hence providing a more attractive service>, a laughing matter. The very concept that passengers ought to or would choose to huddle in a fenced-off subterranean “waiting area” when a station is sited in a CBD is crazy — as is the idea that trains serving the station would run so Amtrak/Caltrain-style 1950s-era infrequently that passengers would wait for large fractions of an hour. If you look a well-designed contemporary passenger rail facilities they’re surrounded by things called “cafes” and “restaurants” and “shops” which far better and far more attractively serve as “waiting areas”, and nobody lurks around stations for hours in advance waiting for a train to arrive.
There is simply no need for a mezzanine of any type, other than to provide bridges between platform-intermediate and intermediate-street levels; the purpose should be solely to redistribute people from any of the (numerous!) street access points to the correct platform for particular trains. Otherwise the space should be as open and airy and inviting as possible — as it is in all contemporary train stations in which actual professionals and competent architects have been involved. Nobody wants to spend time underground in San Francisco — the purpose of the space between the streets and the trains is to facilitate fast and convenient movement of humans, not to trap them in holding pens and herd them through security barriers and turnstiles.
* The TJPA plan funnels all passengers from all points along the length of all trains — which are inherently long and skinny objects — through a tiny knot-hole of vertical access from the underground mezzanine to the surface. Trains are up to 400m long, yet all the routes to the ground level — which is where actual humans wish to go — is compressed into less than 150m. Even better, much of this is in a “great hall” in the First/Fremont block, from which access to and from Market Street (the main real-life human destination) is blocked by the cheesily-copied-straight-from-Hong-Kong office tower and by the traffic of Mission Street.
In a real passenger-oriented facility designed for maximum passenger convenience and maximum throughput, the underground level would be accessible at numerous locations along its length, and the access points would extend laterally (away from the long axis of the trains and building) away from the mezzanine, extending the reach of the building into the surrounding streets, shortening pedestrian routes, and making use of the inherent “reach” of stairs and escalators as they extend. Short, strategic walkways could and should extend to the north side of Mission Street, with access points at Ecker and at the NW corner of Fremont.)
But instead all TJPA vertical circulation lies within the footprint of the building site (the one with the lovely 24’+52’+24′ x 42’6″ grid laid down by God sometime around 2003 and etched in stone since then.) The result is that the structure itself is almost filled by the space required by stair and escalators (what a deal for two billion!) while passenger movement is impeded and pedestrian access times are maximized.
Astrounding, but true!
All in all, the TJPA “plans” minimize train throughput by erecting third-world-style barriers to train movement (tight curves, lack of parallel routes, lack of platform-use flexibility, inflexible Mission Bay station design) and prison-style barriers to passenger movement (platform space constricted and safety compromised by structural column mess and consequently badly-placed vertical circulation, inadequate mezzanine-street circulation, fare gates, cavity searches, etc.)
And the CHSRA “requirements’ exhibit a similar breath-taking combination of incompetence, ignorance and hubris.
Getting 12tph (an arrival and departure every 5 minutes on average) at the Transbay Terminal (peak 3-4 — NOT 12 — HSR, and peak 8-9 Caltrain) or even a borderline-feasible 15tph would be a difficult but achievable goal given first-world design and first-world operating discipline. But given the outright infrastructure sabotage of the rail design, the proposed 19th century operating practices (park your trains here for as long as you like!) and the rank corruption (Los Banos here we come!) of the CHSRA, not even that is possible, let alone some insane fantasy of one high speed train departing every five minutes
Everybody involved is so profoundly incompetent that they could only possibly hold jobs in US special needs sheltered workshop.
PS If CHSRA wants two or three or more deep underground levels of train platforms … umm… how to the passengers get to and from them? Details, details…
Another tunnel under the bay isn’t going to happen.
The TJPA explicitly rejected the only design that would allow this.
(The so-called Second-Mission alignment that would have placed the rail station on a diagonal headed out MIssion Street. Oh, and provided more TTT station platforms and higher train capacity. Let’s not even mention that, shall we?)
Whether this was because of the habitual incompetence of the agency staff and consultants or whether it was the habitual corruption of San Francisco politics (placing the immediate desires of the 301 Mission developer ahead of the 20-100 year needs of the city and region as a whole) is unknown, and probably unknowable.
All we know is that they refused to do any sort of professional due diligence, and went ahead with a design whose glaring deficiencies have been advertised for a decade.
Anway, get used to riding the buses across the bridge (hardly a fate worse than death, you know), because there’s nowhere remotely useful to build a new rail line connection — BART or otherwise — into San Francisco. Think about it! (And no, MTC’s “regional rail plan” did not involve any thinking…)
We had our chance, and MTC/Caltrain/BART/TJPA/C&CSF deliberately and knowingly blew it. Too bad. We have to live with it now.
First off… 12 tph is a completely bogus number. No way is that going to happen, pretty much ever. Among other things, HSRA is seriously expecting people to take the train from SF to Sacramento via Gilroy, which just isn’t going to happen. And I do think that 4 tph is about the highest peak service I’d expect from the HSR for quite a while, As for Caltrain, about 8 tph seems reasonable, with maybe another 2 tph of diesel trains (Gilroy and Dumbarton service) going to the surface level station at 4th/King, which could be reduced to half its present size.
Aside from that, yes, basically what Richard said, only I think that in the 1950s, railroad designers in this country were actually more competent than now, and even Caltrain can do some things right sometimes (the current layout of CP Common isn’t too bad, for example). Oh and I’m pretty sure that NJT and even Amtrak realize that NY Penn is an example of how not to do things, and it’s been a long costly process of fixing up the station layout to fix the passenger traffic flows. Oh and incompatible HSR and Caltrain platforms are a completely insane idea. What are they going to do at Millbrae and Palo Alto/Redwood City?
anonymouse: fully agreed. The CHSRA’s newest service “plan,” or whatever you want to call it, is fantastical, and I’m unclear what sort of justification they can offer other than the theoretical idea that it will be obsolete in five years, that Mehdi’s has been apparently feeding to the Chron. As for the platform issue… yeah, I’ve been wondering the same thing.
I’m no expert but wouldn’t it make more sense to transition Caltrain to the same type of trains used in HSR so that either trains could use any of the proposed six tacks for arrival/departure?
Do you have a high res version of that image?
missiondweller: Caltrain and HSR serve different purposes, travel vastly different distances, have different stop frequencies, and will be attaining different maximum speeds. High speed trains will attain a maximum of about 354 km/hr (220 mph), but in connection with electrification, Caltrain will seek FRA permission to use non-compliant Siemens Desiro EMUs, like this one, which attains a maximum of 140 km/hr. Caltrain rolling stock obviously wouldn’t be suitable for high-speed rail purposes, but there’s no need to use high-speed trains for Caltrain, either, since it stops frequently in a relatively short distance. These planning issues, e.g. platforms, operating plans, etc. are going to have to get worked out. So far, I haven’t been impressed with CAHSR/Caltrain coordination based on what I’ve seen, but there’s going to have to be substantial cooperation in order for the various rail demands of the Peninsula to be met.
egoldin: You know, I could’ve sworn that I did, but I can’t find it now. It was actually larger, and I cropped it so that it would fit into the post. Someone else asked me about this too, so I will keep looking. If I find it, should I send it over, using the email address you provided when leaving the comment? (It appears that the original from Pelli’s website has been taken down.)
Caltrain and HSR serve different purposes, travel vastly different distances, have different stop frequencies, and will be attaining different maximum speeds. Caltrain rolling stock obviously wouldn’t be suitable for high-speed rail purposes, but there’s no need to use high-speed trains for Caltrain, either, since it stops frequently in a relatively short distance.
For the Caltrain portion of the route, the Caltrain EMU and the HSR EMU are functionally equivalent in every possible way. They will have same max speed (125mph), and even similar stopping patterns (i.e. not much difference between Caltrain Express and HSR). They certainly will have same platform height. So, to segregate tracks and platforms exclusively for HSR doesn’t make any sense from a cost or operational standpoint.
Bikerider: oh, I quite agree, it doesn’t make any sense to use separate platforms. And the services are in some sense functionally equivalent within our metro area. (Once Caltrain electrifies, it will make additional stops, and HSR can emerge as the new regional express.) But greater integration of those services operationally — and even same maximum speeds within the JPB corridor — doesn’t imply that Caltrain and HSR will be using the same trainsets (as missiondweller asked), because that question speaks to the entire route, not just a portion of it.
On Caltrain/HSR segregation … there’s a physical problem with the TBT train box design, that the platform tracks at the extreme ends are curved in a way that does not accommodate 16 car HSR trains.
On the trains per hour … 12tph makes no sense. It seems to be people with little operational understanding of high speed inter-regional rail transport throwing together a confused mish mash of airline scheduling and commuter rail scheduling.
Given the planned lifetime of the train box, 8 HSR trains per hour does need to be planned for … 1 LA/Anaheim Express, one Local, 1 LA/SD Express, 1 Semi-Express gets you to four. Add a Sacramento service, specials, the obvious Mojave junction to get to Vegas … 4 trains per hour on the two main routes and another two to four on secondary routes seems quite sensible to plan for.
But there’s no problem using four platforms to cater to 8 trains per hour, so the idea that more platforms are needed is silly.
The problem is the access to the lines. While there are some switches, the current design is basically three bi-directional tracks, one per island, which means each pair of passenger arrival/departure areas share a single track.
4 trains per island means 8 train MOVEMENTS per island, which means each train needs to pass through the tunnel in 4 and a half minutes, safety margin included. And the tunnel is not designed to allow the trains to run at a reasonable speed.
That mess needs to be replaced with an access track, an egress track, and a system to connect to the platforms that reduces the interference between different services. There is, after all, no problem running trains separated at 3 minutes each if they are passing through the tunnel at 30mph, and 3 minutes each means 20 trains per hour, enough for 8 HSR trains per hour and up to 12 Caltrains services per hour during the morning and evening commuter peak.
That saved money on tunneling down 2nd Street (2 wide instead of 3 wide), and puts the same money into a better platform access system with one line passing beneath the rest of the tracks to avoid crossing lines unnecessarily.
Also with 2 access tracks, it is easier to design “second half of the 20th century” curves to allow the trains to run with normal station entry and exit speeds instead of creeping along to get around turns.
As a side benefit, the Caltrains services, which as regional rail transport would have much shorter platform dwell times, gain what they need, which is the flexibility to depart a train from Platform Line 1 while a train is arrive at Platform Line 6 without interference from the HSR services.
BruceMcF: My apologies that I don’t have time to reply more completely just right now, but a couple quick notes:
–I believe that the CHSRA is requesting a redesign of the curved track problem. Not sure how far along that is, or what, if anything, has come out of it.
–I agree with your comments re: 12 vs. 8 tph, and as you remarked, the current platform supply is sufficient to accommodate 8 tph, particularly if you took some measures to increase turnaround efficiency. The proposed third level of six HSR tracks would add unnecessary delay and would be a waste of at least $1 billion.
Oh, and I should probably add, re: Caltrain/HSR segregation. My comment above to bikerider was more directly related to the Peninsula HSR stops, Millbrae and RWC/PA, at least pending some sort of rational redesign of Transbay.
The problem is the access to the lines. While there are some switches, the current design is basically three bi-directional tracks, one per island, which means each pair of passenger arrival/departure areas share a single track.
Hard to imagine.
BayRail Alliance general meeting
SF Caltrain Downtown Extension/Transbay Terminal Update
Thursday, March 19, 2009
Staff from the Transbay Joint Powers Authority will provide an update on the progress of the program to extend the Caltrain rail alignment for Caltrain and California High Speed Rail to a new Transbay Transit Center in downtown San Francisco. Joining us will be Robert Beck and Bradford Townsend.
6 PM – 6:45 PM order your own meal, cafeteria style, at Panera Bread. We will be seated in the back area.
6:45 PM Program starts
301 King Street (across the street from the 4th & King Caltrain stop, next to Muni light rail)
Hi Andy, thanks for posting that notice about the meeting here, I hope to be able to attend.
@Eric, that’d be awesome! Thanks.
Wow. I really hope things get fixed before they break ground!!!
i am retired railroad and an ad hoc Cal HSR advisor.
i recommended a few years ago to HSR project a further revenue contributor and added service to the Los Angeles/San Francisco route. They were:
1) a rail/auto service as Amtrak has between Lorton Va and Sanford Fl that allows passengers to take their vehicles with them in each direction so these passengers have their own vehicles with them at arrival at either San Francisco or Los Angeles. The benefits are several among them no use of petroleum fuels or immissions from driving.
2) Allow HSR to transport trailers or containers on HSR trains between the points served.
this would take diesel trucks off the highway between the points, reducing immissions and also use of fuel.
It would also increase service for companies that needed products at the opposite points in far less time than by traditional highway transit.
Both of these projects would make it necessary to alter the main station plans for Los Angeles Union Station and the Transbay Terminal in the City or outlying stations closeby which would probably be more doable and functional using less valuable land and development.