27 October 2013

Census Driven Service Planning

The diagram at right shows where people live and work along the entire length of the rail corridor from San Francisco to Gilroy, as extracted from government census population and jobs data sets.  The diagram is also available as a PDF.

Implications for Rail Service Patterns

By simple observation of the features of the census population and job distributions along the peninsula corridor, it is possible to infer the desirable features of train service patterns that will maximize commute ridership.
  1. Transbay has more than 100,000 jobs within a half-mile radius (more than every other station in the system combined). The concentration of jobs near San Francisco's Transbay Transit Center cannot be understated. This station absolutely must be served by each and every train, and it would be highly counter-productive to terminate any train at 4th and King.
  2. Silicon Valley shows up in the jobs distribution as a broad hump, mostly homogeneous and stretching from Palo Alto to San Jose.  To serve this rich but diffuse commute market, all trains should make every stop in Santa Clara County.  There should never be any skip-stop service here, and the wider spacing of stops (relative to San Mateo County) will result in only minor trip time penalties.
  3. San Mateo County has numerous stops, spaced more closely together and with middling jobs and population density.  To enable faster service to and from San Francisco and Silicon Valley, it makes the most sense in this portion of the corridor to operate skip-stop express service alongside local service.
  4. South San Jose, while south of Silicon Valley, has a massive and untapped residential market that can serve as origin to jobs further north.  Tamien currently functions as a slow and infrequent addendum to the peninsula service, but should be sped up and extended to Blossom Hill.
  5. Oakdale in San Francisco opens up a new residential market for Caltrain.  The distribution of people nearby is even denser than at 22nd Street.
  6. The Gilroy extension doesn't make much sense.  There are so few jobs and people here that Caltrain (as primarily a commuter service) should not run to this area.  Serving Morgan Hill and Gilroy is best left for a long-distance operator such as Amtrak California.
Putting all this together, what would a service pattern look like that is tailored to the census distributions, to maximize commuter ridership?  First, it would not look like today's Baby Bullet, which severely under-serves numerous stops.  It would also look very different from Caltrain's latest planning fad, the "peak-period skip-stop zone express," used as the basis for all operational simulations in the blended service analyses published to date.  It would look like this, with distances drawn to scale:

  • The Silicon Valley express links the major employment centers of San Francisco and Silicon Valley, highlighted in orange, providing a faster and better alternative to fleets of white luxury buses stuck in traffic on US 101.  It would run every 15 minutes.
  • The San Mateo local serves all the minor stops throughout San Mateo County, terminating and originating across the platform from the Silicon Valley express at Redwood City.  This provides fast and penalty-free transfers between Silicon Valley and cities all along the peninsula.  The local turns back in Redwood City, minimizing crew and fleet requirements while still providing service every 30 minutes.
  • Stops in San Francisco and San Jose that have very large residential markets are served in the peak commute direction only.
This service pattern is also well-suited to future blended HSR service: a southbound high-speed train and a closely-following Silicon Valley Express can overtake and catch up (respectively) with a San Mateo local by using a four-track mid-line overtake facility constructed from San Mateo to Redwood City.  Because the local turns back at Redwood City, it does not impede traffic in Santa Clara County.

The time has come to fundamentally rethink peninsula rail service patterns.  Caltrain's "peak-period skip-stop zone express" is almost certainly not the best solution for meeting future demand; a much wider range of options must be considered.


  1. How does shuttle ridership fit into this. Some stations are just more accessible than others. For example, Mountain View is easier to reach vai 237 than San Antonio would be. If I'm on the east side of 101 in San Mateo County, I'd be running my shuttle to San Carlos via Holly Sr instead of downtown Redwood City.

    Is there data available to plot the shuttle ridership for each station?

  2. For this to work you'd need better transit connections at Cal Ave and especially Lawrence, which has zero VTA connections.

    1. Lawrence is not terrible with the Caltrain shuttles. They could cover more area, but the ridership on them, at least when I rode them, was fairly high. Many could often be standing room only near the station.

      I agree though. Better transit connections would make a lot of these stations much better.

    2. The real issue with Lawrence is that 1) Sunnyvale has stolen much of its ridership north due to superior service and lower fares and 2) limited reverse peak service has hindered ridership growth south of Mountain View.

  3. Excellent idea, though perhaps 20 minute frequencies for both elements? Santa Clara to/from SF every 20 is nothing to be sneezed at, and 20 minutes for the SM local might grow this market.

    1. I think 20 minute service might be a little thin for Santa Clara county, although it does simplify the problem quite a bit since there is 33% more room for trains to catch up with each other, requiring far less overtake trackage.

    2. Number of crews and length of train consist have to be consider at the same time.
      I am afraid some complain from San Mateo area for 30 min frequency while north of RWC have 15 min.
      Differentiate train consist (for example, 6-car SV express, 3-car SM local) and flexible staffing (1 conductor for 3-car train) makes both 15 min frequency easier. In addion, this provide 100% express to local (or vice versa) connection.
      Another advantage of shorter local train consist will be (1) Faster accerelation. (2) Cheaper construction cost for hold-out station both Broadway and South San Francisco.

  4. Clem, Excellent post and this really makes sense! Actually, current limited train which become local south of Redwood City are quite popular and ranks one of highest ridership. This schedule can be fit into current Caltrain's equipment (fleet) and operating budget. How easy to make Redwood city station?

  5. As a San Mateo resident, I have to say that downtown San Mateo is just as important as Hillsdale (Which is my closest staiton), and SM benefits from the current stopping pattern, which provides a 23 minute ride to San Francisco. It's very much a commuter station, as the census shows the large number of residents within two miles. It also has easy freeway access, and good nearby parking. I am no expert, but it would seem to me that retaining a commute and counter-commute type schedule for stations north of RWC might better serve the ridership, as all of these cities are much more "bedroom communities" for San Francisco than the silicon valley cities are. Just a thought - love the idea overall, though! Very novel, and very sensible.

    1. I suspect Clem used Hillsdale rather than San Mateo as the third San Mateo stop for the limited as it's in the mid-corridor overtake section planned by Caltrain, so will have four tracks and could easily be given four platforms for cross-platform transfers. Due to corridor width constraints San Mateo will only ever have two tracks and two platforms under any realistic scenario.

    2. With 15 min express frequency, how far the local train can escape from express? Redwood city need one island platform and 2 tracks both NB and SB and there need enough short signal placing. Then, local train can depart 2~3 min after express departure. With 2 min penalty with each stop, local train may escape from express until Milbrae. There is unutlized NB siding and enough space for additional platform for southbound, it is very easy to built there.
      Provide local to express transfer at Milbrae will benefit to San Mateo and Burlingame because traveling time to SF is very similar to current express train.
      If we use Milbrae as transfer point between local to express, we don't need to build additonal 4 track section until HSR is coming.
      Electrification and lebel boarding will bring higher accereleation and short dwell time. So, the difference between local and express will be shorter. Then, express train frequency can be increase to every 12 min or 10 min but still local train can escape from express until Milbrae.

    3. With 4tph Redwood Shuttle + 4tph Silicon Valley Limited and 0tph HSR there is no overtaking. None!
      Just "catching up" Belmont-SanCarlos-RedwoodCity.

      With 4tph Shuttle + 4tph Limited + 4tph Altamont HSR the only overtake (HSR/Shuttle) is Bayshore-SouthSF-SanBruno.

      With 2tph Shuttle + 4tph Limited + 2tph Altamont (very realistic medium term phase) there is no overtaking!

      But HSR via Gilroy is more complicated and expensive.
      Overtake of HSR/Limited somewhere in MenloPark—PaloAlto—MountainView must happen for any level of HSR. Ouch!
      Also either 4tph HSR or 4tph Shuttle requires second overtake (HSR/local) around SouthSF.

      Not sending HSR trains off Caltrain As Soon As Possible (off to Altamont) is crazy!

      Experiment with http://mly.users.sonic.net/Caltrain-Timetabling/201105-takt/takt.php to see how it works.
      Try out depart from Transbay like Shuttle :00, Limited :11, HS :22.
      All depending on 79mph/100mph/110mph/125mph, it comes out like arriving RedwoodCity Shuttle :45, Limited :46, HS :50.

      All assumes controlled dwell times. Level boarding.
      Without level boarding, give up now and plan for BART extension.

      I don't see how you make Millbrae a transfer station.

    4. There is empty space west side of SB caltrain platform. Add one track for local train and build crossing for outside acesss. Bus and Taxi bay need some modificaiton.
      Local train arrive from San Bruno, 1~2 min before arrive express train. Express train arrives current SB track. 1~2 min after depart express (depending on signal clearance) local train depart to Burlingame. Northbound, use platform 4B. Local train arrives 2~3 min before arrive express ar platfrom 4B. Then passenger walk to platfrom 4. Then express train arrives and 1~2 min after depart express and signal clearance, local train depart to San Bruno.
      I believe this is cheap solution to provide local-express transfer without building long 4-track segment. In addition, access to South SanFranciso will be improve from express train transfer. This will make transfer from shuttle bus to Caltrain.

    5. Here is example schedule both local and express runs every 15 min and follows current Caltrain time table pattern.
      SF Milbrae Redwood City
      5:00----------5:15 5:16 -----------(1 stop)-------------------------- 5:29 5:30 ----- Local to SJ
      5:03----------5:26 5:33 -----------(Local)--------------------------- 5:55 5:57 ----RW Junction sinding (Turn around)
      5:15----------5:30 5:31 -----------(1 stop)-------------------------- 5:44 5:45 ----- Local to SJ
      5:18----------5:41 5:48 -----------(Local)--------------------------- 6:10 6:12 ----RW Junction sinding (Turn around)
      5:30----------5:45 5:46 -----------(1 stop)-------------------------- 5:59 6:00 ----- Local to SJ

      Local train should be 2~3 car consist to get better accerelation. (And 1 condutor!)

    6. Reformatted.

      E 5:00----------5:15 5:16 -----------(1 stop)---------------------- 5:29 5:30 ----- to SJ
      L 5:03----------5:26 5:33 -----------(Local)----------------------- 5:55 5:57 ----sinding
      E 5:15----------5:30 5:31 -----------(1 stop)---------------------- 5:44 5:45 ----- to SJ
      L 5:18----------5:41 5:48 -----------(Local)----------------------- 6:10 6:12 ----sinding
      E 5:30----------5:45 5:46 -----------(1 stop)---------------------- 5:59 6:00 ----- to SJ

    7. Anonymous Millbrae fan,

      I coded up your suggestion.

      As you can see, it requires a lot of new overtaking tracks in the South SF through Burlingame stretch, as well as reconstructing Millbrae station.

      It gets seriously messy when HSR is added (via Los Banos, but everything gets really bad and really expensive without Altamont HSR.)

      But it is bad even with Altamont HS.

      Caltrain Shuttle/Limited timed transfer at Millbrae just looks like a bad idea.

    8. * Here's Redwood Shuttle 2tph + SV Limited 4tph +0tph HSR
      Only infrastructure needed is triplication/quadruplication Belmont-Redwood City and a three-track two-platform (central turnback track with two platform faces; or could be four tracks four platform faces less efficiently) Redwood City station rebuild.
      No overtakes!

      * Exactly the same infrastructure, but more trains and crews, for Redwood Shuttle 4tph + SV Limited 4tph +0tph HSR

      * Exactly the same infrastructure for Redwood Shuttle 2tph + SV Limited 4tph +2tph Altamont HSR. Very very cool.

      * Passing section Bayshore—Millbrae needed for Redwood Shuttle 4tph + SV Limited 4tph +2/4tph Altamont HSR.

    9. In contrast, with insane Los Banos HSR, it's much more expensive and messy ... as well as much reliable due to more overtakes. Win-win-win!

      * An additional passing section is needed somewhere (pick your poison, but change also affects Bayshore—San Mateo ... try changing the parameters, all the alternatives are bad...) in the Menlo Park—Palo Alto—Mountain View stretch for the crazy Los Banos HSR routing: Redwood Shuttle 4tph + SV Limited 4tph +2/4tph Los Banos HSR.

      "Doctor, it hurts when I do this!"
      "So don't do that!"

    10. Richard,
      I need to explain more detail of this proposal. This is 0mph overtake.
      Arrive local train 2 min before arrive the express and depart train 2 min after express departure. So, local train stop at Milbrae 6~7 min. This is slightly longer than current Bayshore/Lawrence overtake but this long waiting will be offset by convenience of express-local transfer. 4 tracking is needed only station area. So, capital investment is less than 500m of additional tracks, two cross over both north / south of station and additional signals.
      There are many example in Japanese commuter train in Tokyo, Nagoya and Osaka area where local to express cross platform transfer is very common.

    11. Richard,
      My proposal intended for current Caltrain schedule and I did not consider HSR blended system. My hope is Caltrain to build enough ridership to justify 4 track system after electrification. By the time of HSR constrution, we should build as long 4 track as. However, design Caltrain schedule both local and express under 2 main track (+2 in selected station) will make HSR build easier. We may need to build 2 HSR tracks at least between south of Bayshore to Burlingame, south of San Mateo to Redwood junction.

    12. The required fleet size for the 4 + 2 + 0 scenario is amazingly small: just 16 trains plus spares.

    13. The Millbrae overtake has the added disadvantage of denying useful service from all of the local stops from Millbrae to Redwood City to all stops south of Redwood City. A lot of time (at least 5 minutes) spent waiting at RWC for the next express. If the trains meet at RWC instead the transfer is crods-platform.

    14. Joey,
      This is proposal for immediate implementation with current diesel trainset and spending minimum amount of capital investment.
      Question is how far local train escape from express? If they can escape until Bayshore, no Milbrae overtake is needed. However, given 2 min penalty in each stop, local train is caught up by express before Bayshore.
      Extending 4 track section further north requres huge capital investment.
      Milbrae overtake will bring new opportunities of faster trip from local station such as Burlingame, Hayward Park and Belmont. In addition, we can send more passenger to South San Francisco who is coming Milbrae by express train. (Those who use shuttle bus from Milbrae currently)

    15. "The required fleet size for the 4 + 2 + 0 scenario is amazingly small: just 16 trains plus spares."
      Indeed. That's the magic of the combination of a service plan, the appropriate rolling stock and strategically targeted investments in infrastructure.

      It's a combination of controlled, shorter dwell times (level boarding platforms and less awful rolling stock), higher acceleration and braking rates (less awful rolling stock), tightened turnback times (less awful rolling stock, marginally less awful operating slackness, and all predicated on controlled dwell times), that enables and is enabled by a simple, frequent, passenger-friendly, resource-efficient service plan.

      Rocket science!

      Play around with the simulated timetables I linked to and try to make anything as simple and nice work with worse performing trains and longer dwells. All the pieces need to play together.

      And imagine what you could deliver on top of this if each of those 16+epsilon trains were operated by a single feather-bed-swadded button-pusher instead of by three or four.

    16. Anonymous Millbrae Fan,

      Having nothing better to do in my life, I coded up your suggestion of a long Millbrae overtake and everything otherwise operated in Caltrain Olde Tyme Commuter Railroading style ("current diesel trainset and spending minimum amount of capital investment.")

      As you can see, the combination of bad train performance, lots of timetable padding (no level boarding), long station dwells (no level boarding) results in both your Millbrae overtake and the Redwood City terminate/transfer catch-up.

      If you want to correct this or come up with something better, feel free. I'm done.

      It all goes to show that getting things to work acceptably really does require an integrated system of Timetable, Rolling Stock and Infrastructure. In other words the exact opposite of the Caltrain approach, which is to randomly spend billions of public dollars (Grade separation! Electrification! Lots more platforms in SJ! CBOSS! Forget Transbay! Forget grade separation! Fourth and King Terminal! Coast Daylight! Blending! More studies! Transbay! More money! Trust Us!) with no plan whatsoever and less than zero attention to delivery of train service to the tax-paying public.

  6. Speaking of which ...
    http://www.sfmta.com/calendar/alerts/5l-pilot-launches-oct-28-5-fulton-changes (diagram)
    * Local short-turns and serves all stops on the CBD adjacent half of the route;
    * Limited is all-stops on the outer half;
    * Limited makes selected major stops in the inner half;
    * Both limited and local are all-stops in the CBD.
    * Pointless and/or too-close stops rationalized


    1. First impression - very cool. I rode the 5L this morning from 18th Ave to Powell. While it didn't feel miraculously different, skipping stops felt like zipping along early Sunday morning with people boarding/alighting infrequently (except there were a lot of people at the express stops and the bus was comfortably full with a small handful of standees)
      We came within about 3 bus lengths of a local 5 hybrid bus, but never did catch it. Whether because the ridership is sorted out better now, or there are simply more buses than before on the inner section, both the local and the limited seemed to make good progress.

  7. This is a fantastic service plan. I knocked up an A/B/C service pattern a while back that gave a high score in your service pattern generator. This service plan achieves something similar with a much more efficient use of resources and a more logical division between zones - 'Silicon Valley Express' vs. 'San Mateo Local'.

    Am I the only person who hates skip-stop service? Either make all stops in a given 'zone' of the line, or operate as a genuine express service through that 'zone' with very few stops. The pain of figuring out whether a given stop will be made by a given service, and if the decrease in travel time is worth the extra time spent waiting for the local, is a real pain even for the transit literate. Corridor legibility becomes more important if Caltrain wants to look more like rapid transit and less like commuter rail.

  8. The data shown is incomplete. Simply saying "there are a lot of residents near Mountain View Caltrain" does not imply that those residents work in SF, in fact since there are a lot of jobs in MV, there is less incentive to commute the longer distance to SF. Residents in San Mateo County however, with fewer jobs, commute to those jobs. The proof is in the pudding - ridership at stations. With San Antonio blocked by Central Expressway and no decent crossover, and Santa Clara blocked out by the airport, those stations will never see the ridership that a station like San Mateo does.

    1. Super insightful. Not.

      That aside, Santa Clara was (pre "Baby Bullet", and could readily be so again) a significant (by Silicon Valley Sprawl Hell standards) traffic generator for Caltrain, just as California Avenue (blocked by Alma ... and with "no decent crossover", thanks to Caltrain's Bob Doty) was and would, without any doubt, be so again with a better timetable and appropriate connecting transit.

      The "pudding proof" you can look at is 101 and 280 traffic and regional commuting surveys, not at the nearly-random set of parking-lot-occupancy-determined hourly train stops made by Caltrain today.

    2. If you see Caltrain's time table 2000~2003, most of train stopped Santa Clara. Its ridership was over 1000 with Feb-00 and Feb-01 ridership counts. (Lawrence also reached 1300 - Feb-01)
      You can find document here

    3. This is the graph you're looking for: winners and losers. The ridership at Lawrence, Santa Clara, San Antonio and California Ave was decimated by the Baby Bullet. It's funny how that happens: all other things being equal, crappy service begets crappy ridership.

  9. It was pointed out to me that if you squint hard enough, South San Francisco looks a lot like Palo Alto as far as jobs are concerned. One look at Google satellite imagery explains why: office blocks as far as the eye can see. Should it also be a stop on the Silicon Valley express?

    1. It depends on the end-to-end travel time. Being able to do end-to-end in an hour minus turnaround time is too useful to give up, although if the trains go to Tamien then the next target, 1:15 minus turnaround, is probably fully compatible with the extra stop.

      I'm a little hesitant about this idea, but it sounds reasonable that if infrastructure investments are based on an integrated takt with timed transfers and short turnarounds then decisions about stopping patterns should, on the margins, be based on what is required to get one-way travel time to nice numbers like 0:40, 0:55, 1:10, etc.

    2. I hope you are being sarcastic? Like, along the lines of "No Caltrain schedule is complete without stops in Palmdale, Monterey and Las Vegas."

    3. IIRC, there's a lot of high-paying (mostly biomedical) jobs around the area. Currently, one hourly northbound limited-stop train serves S. San Francisco, and the number of passengers that disembark isn't insignificant. That being said, from destinations south, the demand for SF will be greater than that for S. SF by several orders of magnitude.

      One possibility is to have it served by every other express train in the northbound direction only, provided that the takt still remains intact. No need to have reverse-peak express trains stop; those riders should just take the local train.

    4. There's some potential at SSF, but before deciding what services will stop there, I think it's worth considering how much the terrible station access situation can be improved. It doesn't matter how many jobs are near the station if no on can get to them, and SSF's location is uniquely shitty as far as access is concerned.

    5. @amanda: not sarcastic at all. Every once in a while you look at numbers and they tell you something that isn't intuitive. In the census plot, compare the orange hump at SSF with the orange hump at Palo Alto. A while back I made this plot entirely from census data and ridership data, without looking at a single timetable--and it did show that SSF was grossly under-served.

      @Joey, yes, station access sucks but it's a problem that can be solved.

    6. @ Joey / @ Clem
      A link to the SSF station's page at the Caltrain website is below. As you can see the SamTrans routes are "within walking distance" [ROFL]. There are also a couple of shuttle bus routes operated by Commute.org.

      South San Francisco Station

      I've made the trek past the station in order to shop at Lowe's (a hardware super-store). I'm a long-time bus / train rider and that was a very sucky trip. I came down from SF on the 292, bought a shelving unit IIRC, and then humped that heavy box back home. At least I had the good sense to bring some rope with me so that the box was easier to grip. The walk back to the 292 stop made me think of Death Valley, the 292 leg wasn't bad, and I beat the commute crush on the 8X. At least now Lowe's has an outpost where Goodman's Lumber used to be.

    7. Here's an interesting post at Green Caltrain:
      South San Francisco applies for grant funding for pedestrian/bike tunnel to Caltrain (GreenCaltrain.org post, 24 Oct. 2013)

      The city of SSF is working towards breaking some of the barriers that US-101 has imposed between downtown SSF and the eastern neighborhoods. This will include shifting the station south to eliminate the safety block (only one train in the station area at a time due to a crossover) and a pedestrian tunnel (northbound) / direct access plaza (southbound) to make the station more accessible.

      South San Francisco Downtown Station Area Plan (project home page)

      Multi-modal Station Access and Connectivity Memorandum ("Station Access & Connectivity" PDF, "Resources" tab, SSFDowntownPlan.org)

    8. Hilarious. Caltrain had construction plans ready for a new SSF station with an island platform with pedestrian underpass.

      Cost overruns on the catastrophically mis-managed and totally counter-productive San Bruno grade separation sucked away all of its funding.

      An application (for money that doesn't really exist) for a tunnel that may or may not connect to a station that may or may not ever be useful in a location that is, thanks to Caltrain stupidity, mostly over to storing freight cars, is what is left.

      Meanwhile "Green Caltrain" just keeps repeating over and over the polly-anna "just give Caltrain more money and everything will be cool" cheer-leading.

      Level boarding ... any decade now.
      A service plan ... we're working on it and may have a draft for Public Input by 2050, almost certainly.
      Just give Caltrain more money first! Trust us!

  10. How would you pull off Northbound AM-only and southbound PM-only stopping in South SJ? It seems you are trying to reduce costs by cutting down on revenue hours of service, but what about the cost of deadheading to position trains at Blossom Hill in the AM and bring them back in the PM? If you're paying to electrify all the way to Blossom HIll, plus paying to deadhead trains past Capitol to/from Blossom Hill, you might as well just stop them there and capture the (admittedly limited) potential ridership.

    1. I'd work on ensuring Tamien has decent service before extending to South SJ. Tamien already has decent peak ridership given its extremely poor headways, and expanding service there may reduce congestion around Diridon Station. Theoretically, much of South SJ could access Tamien via VTA as well.

    2. Yes it is kind of silly, operationally. Since you have to send the trains down there, so you might as well make the stops. It's just that rush hour will only see passengers on one of the two platforms.

      We should definitely look at making south SJ a more integral part of Caltrain. The current boundary at Tamien is entirely artificial, and the Tamien to SJ run times are agonizingly slow. Electrifying to Blossom Hill would require some sort of an agreement with UPRR. These are all problems that can be solved. Diridon would become a short (30 second) stop without all the pomp and circumstance of "terminating" trains, with only two platform faces required for all Caltrain service.

    3. Caltrain owns the right of way all the way past Tamien to Lick, just north of where the tracks meet Monterey Highway. (This is an artifact of the ancient plan to build a joint Caltrans/Caltrain maintenance center in the Lick Quarry.) Between Diridon and Tamien, Caltrain controls the west track, UPRR the right track. To get to Blossom Hill would probably require purchasing the ROW to that point and confining operations to the single track and potentially adding a third track.

    4. Why conflate this "commuter railroad" (northbound AM, southbound PM) tail of south SJ with the Caltrain dog?

      San José, the Capital of Silicon Valley, is very very big indeed on its own rail systems, that being the route to becoming the Paris oif the western US and all. Why not then operate its own San Jose S-Bahn shuttling Blossom Hill—Capitol—Tamien—San Jose Cahill—Santa Clara—Great America?

      Transfer at some selection of Tamien/Cahill/Santa Clara (depending on timetable/infrastructure) to Caltrian.

      It can be operated by FRA commuter railroad trains (there might even be some surplus stuff available somewhere!), craptastic FRA DMU, non-FRA, light rail, or VTA gadgetbahn, but it would be independent, except for Takt timetable integration, of Caltrain, just as ACE, UPRR, Amtrak and all the other junk absolutely must be.

      If VTA wants to run it commute direction or commute hours only, fine. If VTA wants to run it on 10 minute headways 24 hours a day, go ahead, go for broke.

    5. Because then it would look vaguely like Metro North, Shore Line East and New Haven. That would cause the fabric of the Universe to rend.

    6. This comment has been removed by the author.

    7. Clem: The Capitol’s and Blossom Hill’s 44 and 66 daily boardings appear to be a frail basis for building significant future ridership. These Caltrain stops are now competing with the Santa Teresa VTA Light Line’s Capitol and Blossom Hill Stations. A Tamien Caltrain/Santa Teresa LRT cross platform station would likely provide a more frequent and longer service period to South San Jose than a future electrified Caltrain extension. BART currently manages to impose a $4 surcharge on passengers exiting at their SFO Station without severely reducing their SFO extension ridership. A reliably on-time one-seat-ride from Downtown San Francisco for baggage laden passengers trying to make an airline flight connection could account for the lack of customer resistance to BART’s stiff airport station levy. An obvious approach toward enhancing both Caltrain and CHSR financial viability would be to change the present BART Millbrae− SFO link to a 1.435 m track gauge. This Stephenson Gauge right-of-way extension opens the possibility of splitting Caltrain EMUs from San Jose and CHSR EMUs from the San Joaquin Valley at Millbrae in order to enable a one-seat-ride to the SFO Railway Station. An exceptionably convenient, fast and, reliable CHSR connection to a major international airport could be a persistently strong attraction for a large portion of the San Joaquin Valley’s 2 million residents.
      A section of the Millbrae−SFO BART track-way joins the Millbrae−San Bruno track-way. A common-center-dual-gauge-track would avoid additional tunneling expense. (Note: A conventional railroad’s switch check rail−running rail centers are exactly 4.75” apart. This is exactly one half the 56.− 66” gauge difference between BART and Caltrain gauges; the minimum gauge difference needed for a common center dual gauge track.)
      A persistent political movement to extend BART through a Santa Clara Street tunnel to downtown San Jose might be deflected to a more economically efficient route. A branch from the present BART extension construction effort in Milpitas through the San Jose Golden Triangle major employment area to a cross-platform transfer with Caltrain/CHSR subway station directly beneath the San Jose Airport’s main passenger terminal would be a good start. Separate tracks on different gauges would enable fast running on both track gauges before connecting to a subway below Market Street while approaching a 4 track two platform Downtown San Jose terminal beneath Cesar Chavez Park. A BART train’s 10.6 foot width combined with a Caltrain/CHSR matching train platform door height (42”) plus a common-center-dual-gauge-track would allow an any-train-on-any-track-way flexible train operation at a downtown San Jose Station.
      A rail connection to San Jose Airport’s main terminal featuring a cross-platform-transfer ability to a burgeoning Caltrain/CHSR rail transit corridor could yield conditions enabling a significant f Golden Triangle property valuation increase. These business property owners would have a monetary incentive to form an economic development district. Such a political entity could make-way for a BART extension parallel to but mostly 0.7 mile east of First Street largely across parking lots if BART agreed to serve stations less than one mile apart through the district. A San Jose City Council proposal to re-zone the Golden Triangle area to allow up 100 seventeen story buildings in this area would support an major transit capacity increase.

    8. I know the data don't lie, and all, but I am just not
      seeing it.

      Genuinely perplexed! Are you sure there's no coding error?

      Something to think about is that in general the development along the rest of the Caltrain corridor is a narrow-ish strip between the Bay and the Coast Range, and that most travel is north-south.
      (Which is just why the Caltrain corridor is, by US suburban standards, an ideal transportation corridor and should have been fully built-out to contemporary standards decades ago, and have been continually upgraded to improving standards.)
      But once you hit the Mountain View—Sunnyvale border, the width of the sprawl blows out, with the annexation-mad City of San Jose aspiring to stretch from coast to coast, or at least Cupertino to Milpitas and beyond, and there are large major "east—west" ("logical east-west", ie perpendicular to the Caltrain line) highways and arterials and major "east—west" traffic flows.

      You never claim that your simple (but promising!) census-proportional model is super accurate, but maybe it is that case that, whatever its faults and simplifications, it is roughly comparing like with like (geography and development pattern) between Brisbane and Mountain View, with rapid fall-off of any sort of density more than say 3/4 mile east or west from El Camino Real and the rail line.

      Sunnyvale, Santa Clara and San Jose fit less well into that pattern.

      I wonder how your nice summary graph would look with a weighting for residential population that falls off very rapidly (more rapidly than inverse quadratic) after a half a mile?

  11. Re : Oakdale Stn. in San Francisco
    The following is the SFCTA's page on the proposed new station in the Bayview - Hunter's Point neighborhood :
    Caltrain Oakdale Station Study

  12. Good analysis, but it misses a crucial point regarding peak vs. off-peak service for commuter stations versus all-day activity centers. Stations like Lawrence, San Antonio, and California Ave may merit increased peak service to serve commuters (along with strengthened first/last mile connections), but it makes no sense to stop at these stations off-peak. Off-peak service needs to focus on stations that are mixed use, walkable activity centers like Palo Alto, Redwood City, San Mateo, and Mountain View, whereas single use park and rides or employment stations can probably be served by much lower frequencies.

    1. Off-peak headways are probably going to be 30 minutes at best for the foreseeable future, so there's no need for skipped stations (except the express section in San Mateo County). There is huge potential for ridership to grow not only during peak hours, but all day if Caltrain switches from typical commuter rail mindsets to rapid transit a la BART.

    2. Have you ever even visited California Avenue ("Mayfield"). Because it sounds like you haven't.

      Moreover, what do you hope to gained by skipping 2 or 3 designated "commute only" stations only off peak? (The only legitimate answer is "it brings the round trip time under a 30-minute takt boundary and so saves an entire train and crew" ... which doesn't matter as much off-peak anyway.)

      Oh, and work near Lawrence but need to go home early or late some day? "No trains for you now!" is a good way to rapidly teach people "No trains for me in the future, thanks."

    3. Seconded. Crappy mid-day service is the only reason why I had to stop riding the train to work.

  13. Today's census data is valuable - analysis of post-electrification schedule should also look at changing land use in some of these station areas. San Mateo has new development filling in the Hillsdale-Hayward Park rail corridor with Bay Meadows and a bunch of other developments, SSF and Lawrence both have Station Area Plans and are likely to upzone, etc. Good point above about mixed use areas meriting more all-day service, the analysis should include areas likely to become more mixed-use over time, like Lawrence depending on what happens with the plan.

  14. "Caltrain crisis: too many riders" ... "longer platforms, more trains essential."

    Published Monday, December 12, 2013, by the Palo Alto Daily Post

    Caltrain crisis: too many riders
    Railroad needs to add capacity

    By Breena Kerr
    Daily Post Staff Writer

    Caltrain is rapidly heading toward a "crisis" brought on by ridership overwhelming its capacity to carry passengers, Palo Alto City Councilman Pat Burt warned.

    "There is a very significant conflict going on here that has not been talked about publicly," Burt told his fellow council members last Monday.

    The problem is that Caltrain can't add enough trains to carry all the people who want to use the commuter railroad. And that's an important problem to Palo Alto, which is increasingly relying on Caltrain to help reduce automobile traffic.

    "We plan our transportation demand management around Caltrain and yet it's so limited ... they're capped at six trains per hour in peak hours," Burt said, referring to an agreement Caltrain made with the High-Speed Rail Authority. That agreement says that Caltrain will limit its trains so that high-speed rail can use the same tracks for its trains.

    Caltrain spokesman Mark Simon said the current train system is "straining at the seams," and the evidence is often in station delays and packed rush-hour trains.

    "There's just a lot more people on the trains. During some of the peak hours the trains are above capacity, which means some people have to stand. Trains have to wait longer at the station ... It is a problem and it does affect our ability to maintain an on-time schedule," he said.

    Ridership has risen to 47,060 boardings on weekdays in 2013, up from 42,354 in 2012 and 37,779 in 2011, according to Caltrain statistics.

    Ridership hit a low in 2004 with 23,947 boardings each weekday, according to Caltrain.

    Electrification is the future

    That demand, which is expected to rise, is a big part of the reason why Caltrain wants to switch from diesel to electric trains, which can stop and start faster, allowing for more efficient use of the tracks.

    "We're kind of straining at the seams. We're pretty much at capacity for what the railroad system, as it's currently constructed, will allow ... (with electrification) we can increase service, we can stop more times at the stations," Simon said.

    A Sacramento judge last month blocked the California High-Speed Rail Authority from selling $8 billion in bonds and ordered the agency to rewrite its funding plans. But Simon said that ruling won't stop Caltrain's plans to modernize.

    He said Caltrain already has the money for the first phase of its modernization plan, which includes updating its signal system. The rest it will worry about later.

    New signals and controls

    The first phase includes a new signal system that will allow trains to run more efficiently and closer to one another, as well as fail-safe controls such as those that would automatically prevent trains from speeding.

    "We're not at a point right now where getting the funding makes or breaks what we're trying to do," Simon said. "We have to wait and see how this plays out."

    If high-speed rail does fail, Burt said Caltrain might still find a way to pay for electrification and modernization. But whatever happens with funding, Burt said Caltrain should consider adding cars to its trains and lengthening train station platforms as a way of increasing capacity.

    "I think it is the best way to go, it's not cost free to lengthen the platform and add more trains, but it is essential," Burt said.

    1. Caltrain doesn't do 6 trains an hour yet. So they haven't even maxed out that.

      Even with 10 minute headways, double-decker 6-car trains can handle quite a bit more than 50,000 people a day. The decrepit Orange Line of the MBTA handles close to 200,000 people a day, and it seems to be falling back to 6 trains (of 6 cars each) per hour lately.


    3. Burt said Caltrain should consider adding cars to its trains and lengthening train station platforms as a way of increasing capacity.

      Longer platforms? Good luck with that. Caltrain's station Design Criteria says platforms shall only accommodate 6-car trains.

  15. I don't think Caltrain need to limit 6 train/h. If Caltrain need more train, just build more bypassing track for HSR.

  16. Maybe I'm biased, but I don't see how an Oakdale station would attract that many riders, irrespective of the local density. How many of those people around the Oakdale station would be commuting south? How many are commuting south today?

    That said, the neighborhood could gentrify. Then they might get the traffic.