Skip to content

Recent Comments



    At the board workshop yesterday, the CAHSR rep said that Metrolink is looking to Northern California.



    Will you be funding the 10’s of billions of dollars worth of research, development, testing, and engineering costs? And that’s even before considering Elon’s orders-of-magnitude-too-low construction costs.

    How about just getting Elon – a billionaire – to throw in a billion of his own? According to his made-up cost projections, that’d cover 15% of the cost.



    Apparently you don’t understand that Transbay has more jobs within a 0.5 mile radius than within 0.5 miles of every other Caltrain station COMBINED.



    The fact that you’re touting SEPTA or NJ Transit as examples that Caltrain should strive to follow is evidence enough that your opinion is worthless.



    It makes perfect sense when you consider that stations can’t be adjusted (ie rebuilt) overnight. There needs to be a solution during the interim for dealing with two different boarding heights, and dual-height doors is the best way to address that issue.



    Wow, I find streetsblog’s reporting on transit issues to generally be fact-based and based on international best practices and experience (ie everything that mainstream reporting on transit issues isn’t) but this post is just flat-out sensationalist and misguided.

    Could streetsblog PLEASE reach out to Clem Tillier for a rebuttal, or, better yet, just repost his detailed and fact-based analysis of the issue? (everything that this post by Andrew Boone isn’t)


    Jamison Wieser

    Barcelona has about the simplest solution possible for curb extensions: modular concrete pads, with metal grates bridging slight height changes and to not effect drainage.



    You can’t make massive infrastructure decisions because you ‘suspect’ Caltrain would do something stupid and illogical when operating the extension. You need to do better than that.

    There would be no logic for not stopping a Caltrain at 4th & King if it’s also serving Transbay; it has to pass through the station anyway, and there’s clearly enough demand at the station to warrant it. There is already a bypass track planned at the new underground 4th & King, so stopped trains would not be blocking the mainline.



    It’s an off the shelf train with an extra set of doors; not exactly rocket science. Caltrain themselves said they got a “good response” from car builders when they were asked they to accommodate this.


    Harvey Kahler

    Not a separate terminal for AT&T Park but an intermediate station. I had trouble finding any plan for a Transbay terminal on the Caltrain site. Space would be limited for a two-track terminal, let alone for four or six tracks anywhere near the BART Embarcadero/Ferry Terminal on Market Street.

    2nd & Market would be a better Caltrain/CHSR terminal location for transfers to BART/Muni Montgomery station from their train load perspective as well as being a more central downtown location. A 2nd Street alignment would allow a new intermediate station at 3rd Street for AT&T Park.

    However, 2nd Street may be too narrow for more than two platform tracks. A larger terminal of 4-6 tracks and a tight 400′ radius curve would entail taking property; but this land could be redeveloped.

    I’ll probably be dead before CHSR begins running into San Francisco; but an extension to BART and Muni connections at Market Street would seem to be a more immediate priority. Costly electrification can be held off until then. The Pacheco connection will have to be built first somewhere between Fresno and San Jose to get into San Francisco.

    In the meantime, it always seemed to me that a high speed capable connection is needed most urgently between Sylmar and Bakersfield to begin through service, albeit at more conventional speeds (<110 mph), from San Diego through Los Angeles to Oakland and Sacramento and intermediate stops.


    Harvey Kahler

    I should add that 850′ platforms should be built now at a Transbay stub terminal for an eventual build-up to 10-car trains, even if this is a longer train than in current service. And don’t forget the locomotive.


    Dexter Wong

    But a bus yard must be reasonably close to the lines it serves. Kirkland serves lines in the north and west of the City, while Woods (in Dogpatch) serves south and east of the City. If a new yard is built in the Bayview, then Woods would have to take the place of Kirkland and the new yard would take the place of Woods. There would be a travel penalty for additional time spent getting from Woods to the assigned lines.


    Dexter Wong

    Bus stop bulbouts have been used on Polk Street for more than 40 years. I remember the first one between Jackson and Washington Streets. It seems your idea of auto paradise is a mall parking lot.


    Harvey Kahler

    BART does not have the same problem of dwell time for boarding and alighting as does California High Speed Rail and Caltrain (and Metra and other commuter roads) with a full car loading or unloading at the downtown terminal.

    Boarding ideally takes 3.8 minutes through the doors and up the steps for gallery cars; but that can be hindered by passengers searching for a seat. A train coming in with passengers in the reverse peak direction will add more time. Add 2-3 minutes each to move through the terminal throat and you may only get eight trains an hour with two platform tracks. That results in a nominal Caltrain capacity of about 11,000 passengers in a peak hour, assuming 10-car trains.

    CHSR trains may take a little less longer to board with fewer seats in each car than Caltrain, but carry-on and checked baggage could slow boarding. But if there is only hourly or half-hourly CHSR service, additional capacity could be exploited by Caltrain if equipment is compatible. That could allow another six Caltrains an hour, raising the hourly capacity to nearly 20,000 an hour, the equivalent of 9 expressway lanes in each direction compared to one track each way.

    In reality, a more reliable schedule may be for only ten Caltrains an hour – that’s still as much as 14,000 commuters an hour sharing CHRS platforms. By comparison, a Chicago Transit Authority subway only can handle 12,000 passengers and hour reliably with 8-car trains on 3-minute headways.

    Another problem is that you can have two streams of passengers moving in opposite directions on the platform hindered by those who came earlier and are waiting to board; so platforms need to be wide enough. One way around a limited platform width is to have additional stairs or escalators along the platform leading to a mezzanine level or the street.


    david vartanoff

    I do not support mandating a smart phone, but combined w/ a contactless card (Ventra/Clipper etc, and I am adamant that Metra within CTA’s current service boundaries must honor CTA fares for travel) fare collection should be easy. Most of Metra can emulate Caltrain which has TVMs and card readers on all platforms. POP chjeckers read your Clipper card to see if you tagged before boarding, and remind you to tag out on alighting. Asto MED the irony is that the Illinois Central was a “bleeding edge” adopter of mag stripe farecards in the 1960s, but Metra scrapped the system returning to 19th century tickets sold by humans at stations.
    One of the side benefits of employer encouragement of transit use (as well as bulk pass purchase by universities and others) is that a rising percentage of riders are regular pass users which means the single ticket sales are a declining fraction of the riders. This in turn gives the agency cash up front in large deposits and lessens the importance of catching every single scoflaw rider.



    construction costs alone were between $200,000 and $500,000?

    Plus something like this:

    Sure it is a bit more complex than it seems so lets multiply this by 10, that makes about $30.000


    Harvey Kahler

    I think the 25″ high from top of rail platform is considered mid-level and it must be farther from the center of the track than an 8″ TOR platform. At some older stations, the platform can be below the rail after a number of track lifts with ballasting. Even where the floor and platform height match, a bridge plate is needed across the gap for wheelchairs.



    So using Europe’s most popular HSR train is “risky”, but having Caltrain design a special snowflake train isn’t? Well okay then…


    Harvey Kahler

    Modern fare payment technology isn’t always appropriate. Even as Metra is about to roll out smart-phone ticketing, this may slow down on-board fare inspections on an open system for lines other than the Electric District. Needing a smart phone will be an economic deterrent for people with little or no income who may still need to take the train, given the much higher cost of a taxi or downtown parking. Automated fare vending of flash-type tickets is needed at all Metra stations and purchased before boarding.



    During the transition period, the boarding situation proposed above would not be any worse than the current situation.

    This is a multi-billion dollar, multi-decade project that results in more stairs, and more lifts. Surely we can do a lot better than that.


    Captain America

    Cancel this boondoggle and build the Hyperloop up I-5.


    Harvey Kahler

    A veteran bike rider can board about as quickly as someone toting a large roll-on case on Metra which is similar to Caltrain. What takes time are the wheelchair lifts; and the need is irregular on Metra. This will mitigate against faster schedules on closer headways.



    Good question, @BBnet3000:disqus. It is actually much easier to facilitate drainage through parklet design than bus bulb design, as you don’t have to alter the cross slope of the road. Instead, you just build the platform around it, allowing for drainage underneath the parklet. The stoplet would function similarly. You can check out design guidelines in the SF Planning Department’s parklet manual on pages 35-36:


    Harvey Kahler

    Amtrak Surfliners, Capitols, and San Joaquins, Caltrain, and Metrolink have low platform boarding cars; so the boarding compatibility issue goes beyond the San Francisco Peninsula.

    Seats are the normal expectation for suburban service capacity and avoidance of standees. With more doors for faster boarding, you have fewer seats and are hauling doors instead of passengers. Does it help Metrolink and Amtrak to have two double doors on each side of the car? Is the faster boarding due more to queuing on the lower level and a short step off?

    With flash type tickets, the gallery type car offers excellent on-board ticket inspection. Unfortunately, Metra in Chicago has not taken advantage of automated ticket vending or even have agents at all stations, obliging crews to handle cash fares only the old-fashioned way.



    1. Wrong. The FRA hasn’t even released the the rules yet for true HSR (186+ mph) since it doesn’t exist here.

    2. The Eurostar just purchased new train-sets from Siemen’s Velaro family. And guess what, those have 48″ boarding height.

    3. Again, the TGV duplex is not the same train-set as the one which set the speed record. So no, it does NOT go much faster. You need the AGV which has 45″ boarding.


    Andy Chow

    No. I suspect that many trains will skip 4th & King.

    Baseball games create massive demands. If there’s no terminal or turn back track at 4th and King, trains would be blocking the main line while waiting for passengers to board. There needs to be more capacity and flexability as well at 4th & King



    “This may be true for commuter rail systems that always have high platforms and use bi-level cars, that these are real issues when it comes to Caltrain where all stations are low platforms.”

    That makes zero sense. Almost every commuter rail agency in the US has high and low platforms, such as MBTA, NJ Transit etc. At high platforms, wheelchairs roll on. At low, stairs pop out and a conductor brings around a mini lift. It works exactly like MUNI. Boarding from the low level makes zero sense.


    Nicholas Littlejohn

    This makes no sense. We need to compromise or adjust stations, I feel.



    How big of an issue is drainage? There are bus bulbs that still allow waterflow in the existing gutter space:,-73.998811,3a,43.4y,232.3h,81.15t/data=!3m4!1e1!3m2!1sagUx5zDyqWJ07jV8aLQzlg!2e0!6m1!1e1


    Mesozoic Polk

    On the one hand, the Mesozoic Polk Neighborhood Association vigorously opposes any so-called “improvements” that make streets more pleasant or encourage anything other than parked cars to linger by the sidewalk.

    On the other hand, if we absolutely must tolerate this bafflingly popular and resilient parklet program, at least having parklets double up with bus stops would preserve more curb space for its rightful purpose: car storage. (For reference, see Bob Gunderson’s excellent guide to street taxonomy, which explains how street parking is a state of nature.)

    To be honest, we are torn on this one.



    We need a solution that improves things over the longer term, and does not make things any worse during the transition period.

    During the transition period, the boarding situation proposed above would not be any worse than the current situation. Once the platforms are rebuilt we would have level boarding, which would be a major improvement over the current situation.



    If I’m not mistaken, many of the Caltrain cars today have people boarding from 8″ platforms onto 45″ floors, up several steps. Even a 50″ boarding would involve level boarding, then down a step or two (or maybe even a ramp) after you’re inside. It seems like it would be a lot better than the current system. Although I can see why a completely flat lower deck would be preferable.



    “Not worse” isn’t good enough. The goal of level boarding is reduced dwell time (better travel time) and increased reliability (better transfers), so for the money a solution should be better than today.



    No one’s actually waiting for BART in the plazas, obviously, because you’d miss your train.

    However, I’d say parklets have been successful. And there’s no reason why a parklet which is also a bus stop shouldn’t be any less successful. I think the key is the buy-in from neighboring businesses who keep an eye on it and have an incentive to take care of it– just like any other parklet.



    Ha. They’re not thinking about it. As someone else has said, Caltrain are just starting to deal with a problem that Metrolink hasn’t yet realized that they also have.



    im really sure you don’t believe this nonsense.



    these spaces get filled with the homeless and we as a city refuse to deal with the homeless problem.



    “Internal stairs mean cyclists trying to move bikes up and down while train is moving. This is a safety hazard.”

    In your opinion. I don’t see cyclists going up or down 2-3 stairs as any more dangerous than non-cyclists going up or down the much greater number of stairs to reach the higher level of the train cars.

    “As long as FRA requires bike racks rather than letting bikes stow where ever (which means there’s a capacity limit for bikes), there should be dedicated bike cars.”

    Surely this requirement applies to bike cars as well as regular cars? I don’t see why rack space for x number of bikes spread across the entire train would take up any more space than rack space for x number of bikes in a bike car.

    “You cannot have cyclists running up and down the platform to look for room in different cars.”

    As happens all the time on BART… right? Expect that it doesn’t. Sometimes you need to move to the next car to board, but it’s hardly a big deal.

    “You cannot have cyclists to occupy ADA space or conflict with ADA passenger boarding.”

    There’s no reason why this has to occur. Bike space and wheelchair space can and should be separate, regardless of the train design.



    Have you seen the huge numbers of people who pile off the trains at 4th & King every day at rush hour and cycle to the financial district, or take shuttle buses or Muni Metro to the financial district? Not to mention the even bigger numbers of people who don’t take Caltrain because cycling or taking Muni from 4th & King is a PITA?

    Those are the people who make up 90+% of existing and potential Caltrain ridership, and the ones we should be designing the system around.



    Every train that serves Transbay would also stop at 4th & King.

    Having a separate terminal at 4th & King does not increase service to 4th & King, it simply reduces service to Transbay.


    Andy Chow

    Internal stairs mean cyclists trying to move bikes up and down while train is moving. This is a safety hazard.

    As long as FRA requires bike racks rather than letting bikes stow where ever (which means there’s a capacity limit for bikes), there should be dedicated bike cars. You cannot have cyclists running up and down the platform to look for room in different cars. You cannot have cyclists to occupy ADA space or conflict with ADA passenger boarding.



    For sure, it should be tested, but I can’t imagine that it would be any worse than the current situation on the Gallery cars. It means you would have to go up two or three stairs, turn to your left or right, and go up another two or three stairs. Right now, you go up four stairs, turn to your left or right, and walk forward. Not a huge difference.

    I do think that having a dedicated bike car is a bad idea as it causes congestion as cyclists all have to enter/exit at the same location. It would be better to have a dedicated bike space near the doors on every car to spread out the crowds. Dealing with that issue would be far more effective in reducing dwell times.


    Andy Chow

    Not all the SEPTA and NJ Transit stations have high platforms even on the Northeast Corridor. So the idea that somehow all the platforms be converted can be wishful thinking.

    I would also like to know the path for Muni to achieve ADA access at all Muni stops. Who could disagree that all Muni Metro stops should be wheelchair accessible?


    Jamison Wieser

    Caltrain isn’t the only shared corridor with California High-Speed Rail and I’m curious how these situations are being handled by Amtrak and commuter rail lines though LA and San Diego?


    Andy Chow

    Baseball alone is more than enough of a reason to have a separate terminal at 4th & King. Baseball is popular enough for ferries to have a separate ferry landing by the ballpark. Baseball is more than enough for Muni to have a separate terminal at 4th & King with tail tracks rather than sending every train down to the Bay View on 3rd Street.

    People are already complaining about the overcrowding before and after Giants games when Caltrain does not provide extra service.

    If it is all about development, shouldn’t Muni sell the bus yard by Fisherman’s Wharf and build a new yard in the Bay View instead.



    given that Transbay is a much more desirable destination for most riders

    apparently you’re not a giants fan….



    “Why can’t we create a space that people actually want to sit at, that
    would make people excited to wait for a bus?” he said. “Instead of being
    a waiting experience, it can be a relaxing experience.

    Because we’ve tried that and the results are terrible. See: Powell St. station, 16th and 24th St. BART stations, etc. etc.

    I’m all for public places, and I’m all for public transit, but the two just don’t mix well — at least not in SF.



    Is this the case? Caltrain really does serve lots of passengers with bikes as a result of our land use pattern. I don’t think anyone knows the dwell time affect of the number of bikes in a car with internal stairs. Seems like something that should be tested.



    There is no reason to originate or terminate any services at 4th & King, given that Transbay is a much more desirable destination for most riders, and could handle all Caltrain service with a bit of coordination between agencies. Every train that originates or terminates at Transbay would also serve 4th & King/Mission Bay, of course.

    There will need to be a staging/storage location in or near SF, but it doesn’t necessarily need to be at 4th & King. Bayshore would be a perfectly good location, a would the Industrial area near Islais Creek.



    You could store the bikes in the lower level instead, but leave wheelchairs at the middle level. In fact, I think that’s probably a better solution. Again, to do the thought experiment:

    At high platforms: wheelchairs have level boarding at the high door. Bikes enter at the high door then go down a few steps to the lower level, or visa versa to exit the train.

    At existing platforms: wheelchairs use the high door with lifts or mini-high platforms. Bikes enter/exit through the low doors, going up/down a couple of steps as they do so.

    You could also have a bi-level train which has space for bikes in both the middle section and the lower section.

    The point is, any solution should try and avoid making wheelchairs change levels within the train; but making cyclists move been the lower level and mid level (or visa versa) is not a big deal.