19 September 2023

First Look at Electric Service

Caltrain's proposed weekday peak service
Caltrain has started to pull back the curtain on their fall 2024 service pattern, when the electric fleet will (finally!) enter service.

The peak service pattern is depicted at right from a Caltrain slide, and looks like this in a string diagram from our trusty taktulator. The overall score for this timetable is a lukewarm 115 points relative to the benchmark score of 100 for the 2011 timetable, and it requires just 14 trains to operate, not counting spares.

The Good

  • Peak frequency will remain at four trains per hour per direction until ridership recovers more. This requires a waiver from the FTA, which originally made its funding contingent on operating six trains per peak hour per direction.
  • This enables operating with all-electric service in the portion of the corridor between San Francisco and San Jose, relegating the diesel fleet to where it belongs: not under the wire.
  • Stops are restored throughout the corridor, increasing service frequencies especially in Silicon Valley where the 2004 arrival of the Baby Bullet decimated ridership at skipped stops that nevertheless have healthy numbers of nearby jobs and residents.
  • Off-peak service is a subset of peak service, making for a regular and consistent half-hour "takt" throughout the entire day. This makes the timetable easy to use and memorize.
  • The Gilroy branch is served by a cross-platform transfer in San Jose, reflecting the results of a recent survey of its customers that revealed that a one-seat ride was their lowest priority.

The Bad

  • The Baby Bullet unfortunately survives as Express A, skipping perfectly worthy stops throughout Silicon Valley and tragically breaking the potential symmetry of the timetable. Express A should just be another Express B, giving evenly spaced 15-minute service from Redwood City to San Jose. This is very low hanging fruit that improves the timetable score from 115 to 118 points. The Baby Bullet pattern was a great idea for speeding up diesel runs, but in the new electric age, Caltrain should let it slip into history.
  • The "South County Connector" remains as diesel rail service. There is an opportunity for  enormous savings (and better service quality for riders) from selling off the entire diesel fleet and operating the Gilroy branch with much more frequent luxury buses. Operating and maintaining an excessively large electric + diesel fleet and wasting capital funds on the expensive but dubious idea of a BEMU distracts Caltrain from its core mission. If one believes in the idea of a fiscal cliff, then it's obvious that rubber tires can help Caltrain's balance sheet until the high-speed rail project electrifies the south county corridor.
The Implausible
  •  The proposed service pattern is built around a 75-minute all stops local trip between SF and SJ. This 75-minute figure is very impressive and would be a huge strength, if only it didn't border on the implausible. These are several issues with a 75-minute local run:
    • the trains have to accelerate very aggressively, something the equipment is admittedly capable of, but ouch, that electricity bill! Operating cost will depend not just on usage but also peak load, and high acceleration makes the peaks worse.
    • the timetable padding has to be shaved down significantly from today's comfortable margins
    • the station dwell time has to come down to a crisp BART-like duration of just 30 seconds. Without level boarding, this is unlikely to work reliably in daily service. The boarding step arrangement, door spacing and interior circulation of the new EMUs is identical to the Bombardier diesel sets, so it's hard to believe they would board any faster, before even considering the possibility of a crew interventions for wheelchair users.
    • There is nothing physically preventing a 75-minute trip, but if Caltrain can pull that off, what secret sauce of brief dwells have they been holding back from us all these years? Based on past operating practice, a more reasonable expectation would be an 86-minute trip. Let this serve as a prediction that Caltrain will find out the hard way that it really does have an urgent need for level boarding.

Taken together, there's a lot to like in this emerging service pattern, and a few tweaks can make it even more optimal. With freeway traffic getting steadily worse, the new service product will sell itself.


  1. Here's my guess at the 2024 schedule.

    I based my schedule on the northbound BART transfer times at Millbrae given in the service plan, and I ignored their estimates of end to end times. It's interesting to see the shade thrown at BART in the service plan for BART's reduction of Millbrae service. As always, the rivalry between the two Bay Area rail operators sees them pushing blame onto each other and hurting riders of both systems. My question is, why does the new BART schedule not have the yellow line descend to Millbrae, with the red line terminating at SFO? I understand their decision to have all trains go through SFO, as SFO is not only closer to the San Bruno stop but also has higher ridership, and according to BART this change apparently mitigates some supposed "confusion" of SFO riders. However, having Millbrae on yellow instead of red gives both Caltrain transfers and Millbrae originating riders a much smoother experience. Not only would this give the station the 6 TPH common to the rest of the yellow line up until Pittsburg, it would also allow these riders to access the yellow line's improved timed transfers to orange lines going north to Richmond. Currently, these Millbrae riders relegated to red line trains must get on a train at Millbrae, ride for one station, and then wait up to 10 minutes for a yellow line train to make a transfer towards Pittsburg. If the point of the 6TPH service of the yellow line was the extraordinarily high ridership at the yellow line stations, why make Millbrae riders wait 10 minutes at an untimed transfer for service to these stations? While Millbrae may not have the ridership to command direct service to San Bruno by skipping SFO, Millbrae has the most riders of any station relegated to the pitiful 3TPH weekday daytime service. What makes this the most infuriating is seeing the short, one stop distance on the map prohibit Caltrain riders from much quicker transfers as well as smoother timed transfers farther into the BART system. I'm sure theres some issue I'm not aware of that makes it harder to turn these trains around at Millbrae than to turn them around at SFO, but its disheartening to see the losses to Millbrae service for Caltrain riders and Millbrae residents.

  2. Speaking of needlessly running up the power bill and other O&M costs, Caltrain appears set to always and only run 7-car trainsets, regardless of demand. Meanwhile, most of the rest of the world’s EMU operators avoid wastefully running over-long off-peak trains by using shorter sets and running them as longer “Schaku-coupled” nose-to-nose rakes when needed.

    1. It has been shown that switching can become quickly more expensive than just let the full consist run.

      The (I think) biggest operator of KISSes (SBB for S-Bahn Zürich) has 6-car units, and the option to strengthen to two sets. However, this requires according platforms (300 m in this case).

    2. @Max: BART recently announced it expects to save $12m (per year?) while improving rider safety by “right-sizing” (shortening) its trains to more closely match actual passenger loads. Unlike BART, Caltrain requires conductors and so it can realize even more savings (than just saved energy and vehicle miles) by using only 1 conductor per shorter 3- or 4-car off-peak train.

      Also note how many trains shown in this German S-Bahn video obviously use shorter 3- or 4-car sets coupled together. What do you and Caltrain know about quick & easy automated Schaku (un)coupling costs that they all (apparently) don’t?

      It seems Caltrain would be able to run more cost-effective (and therefore more frequent) off-peak (midday, night, weekend, holiday) service using shorter trainsets. Being stuck with all & only 7-car trainsets sized for peak period loads and maximum platform lengths seems like an obvious strategic error that works against providing more and better off-peak service to grow and serve the more important post-pandemic non-commute ridership market.

    3. FWIW, a BART board member commented that splitting trains in the off peak didn't really save them any money. I get the impression that reducing the train size is more to do with the lack of (usable) new train cars which have lower operating costs compared to the older fleet than saving maintenance generally. BART has also been expanding train sizes and running old trains again anyway so I doubt that $12 million reduction gets realized.

      Caltrain would of course be able to reduce personnel costs by splitting trains or reducing the overall length. However, given how unusually configured these KISS trains are it would require a pretty significant change in their current order, likely pushing back electrification even further, which is absolutely one of the biggest ways they can save money in the near term.

    4. @Nick, yes, having/using shorter trainsets means a higher ratio of cab and bathroom cars is needed since every trainset needs 2 cabs and at least one bathroom. So to be filed under “much harder & costlier to fix now.” Caltrain planners (and everyone) should’ve thought more carefully about trainset lengths before ordering all and only 7-car sets. :-(

    5. Caltrain’s getting all 7-car sets is a like an airline replacing its tired old fleet with only A-380s. It guarantees needing to either serve only high demand trips or uneconomically needlessly flying around a shitload of empty seats and paying unneeded flight attendants (conductors) to provide good off-peak / lower-demand service while excessively guzzling away costly jet fuel.

    6. Yeah, there'll be plenty of seats on the 7-car sets. It's likely not great from the operator perspective, but from a passenger side it's fixed sets are just nicer and friendlier. Consider these benefits:
      * Gangways
      --- Gangways between cars are cleaner because they don't need to be weather proof.
      ---Just general travel between cars can be made much better - although I haven't toured the cars yet.
      * Trains with a Scharfenberg coupling don't usually permit travel between cars. Lack of passage leads to...
      --- Less safety, as passengers can't get away from a nuisance (see BART)
      --- Conductors can't assist each other to deal with nuisance if they are away from each other
      --- Both of the above lead to less safety. See where "safety" ranks for keeping riders away from BART
      * Coupling and un-coupling is still a non-revenue operation that costs $$$$, so there need to be more than enough savings to make it worthwhile.

    7. @Reality Check: Yes, German S-Bahn Networks have always operated 70m long trains, with 2 or 3 units coupled together to handle peak loads. But in normal operation, they aren't uncoupled into single sets outside of rush hour, simply because that would require additional crews, tracks, and time. The Munich S-Bahn recently placed an order for their next generation fleet, those will be 210m long trains because shorter trains won't be operated anyway and small, coupled sets simply cost more. The same (or similar) trains will most likely be ordered for the other networks that have the same characteristics (Stuttgart, Frankfurt, Cologne) once they announce their plans for future fleets.

      In other words: There are very few and very specific circumstances under which coupled short trainsets perform better than single large ones. For example, if one fleet is supposed to serve multiple lines with different demand, running single trains on some lines and coupled trains on others makes sense. Also, if long trains would't fit into the maintenance facilities, purchasing shorter trains can be cheaper than rebuilding all that infrastructure.

      But Caltrain faces neither of those challenges. They operate just one line, which has high peak demands, and where splitting the trains only to save some money in between the peaks would probably cost more than just letting the trains run rather empty, something that a lot of rail operators have learned over the past decades.

    8. Shhhh! As explicitly mentioned by Caltrain board & staff at public meetings going back several months now, Caltrain has and is continuing to now run many shorter 4-car trains (instead of all 6-car trains) to … wait for it … save money! ;-)

    9. In-service Hamburg S-Bahn Schaku coupling (and many other video clips like it) … show that it’s obviously super expensive! :-)

    10. It's not about the mechanism, the cost comes from everything around it! In Hamburg, two branches of line S1 meet in Ohlsdorf, where trains are combined for the rest of their trip. This means that there is always an operator for the second trainset, something that is not true for the type of shunting moves that Caltrain would have to do to get half of every local parked outside of peak hours. They would need additional staff just to move those 'dead' sections of trains out of the way, which doesn't help when you're trying to cut costs...

      On top of that, in order to have a schaku coupler in the first place, you'll need two cab cars instead of the 7th power car, which (when using clem's cost figures from the comment section of the BEMU post) makes a 4+4 set 16.6% more expensive than a 7 car set. This means that instead of having 19 trains available for peak sevice, there would be 32 or 33, or just 16 long sets, making more than the 4tph service pattern impossible. In fact, this would force Caltrain to keep even more of their diesel fleet in order to comply with the FRA-mandated 6 tph at some point in the future or purchase additional electric trains - both options increase the overall costs and reduce the savings from short train operations.

    11. Caltrain running 4-car trains vs. 6-car trains differs significantly from what's discussed here.
      This shortening is done at the maintenance facility due to the need to reconnect the locomotive control, power, braking, tighten the coupling, and open up the gangway. Caltrain could take their 7-car EMUs and shorten them to 5-car at the facility. This would be very labor intensive, so it's not something to do as part of the operation.

    12. Caltrain reducing the length of their diesel trains was also super temporary based on my experience. Trains are mostly back up to 5-6 cars at this point. Also, I’m fairly certain the reason they did that is because they didn’t have enough usable trailer coaches for a little while. So many of them were in the shops for repair they went well under their typical spare ratio. “Cost cutting” was simply a convenient excuse for them.

    13. While it’s true that riders cannot move between sets coupled nose-to-nose (and so?), inter-car gangways are uninvolved and unaffected when quickly Schaku (un)coupling 3- and 4-car fixed sets at SF or SJ terminals (or at nearby CEMOF), when needed, to more closely and efficiently match train lengths to runs with peak and off-peak ridership.

      Other configurations are possible, but supposing Caltrain’s 168 cars were a mix of 3- and 4-car sets (instead of 24 7-car sets) short single-set train lengths could be 3 or 4 cars and longer sets could be 6 (3+3), 7 (3+4) or 8 cars (4+4). If Caltrain (or whoever) insisted on keeping conductors instead of replacing them with OPTO combined with professional, dedicated roving fare inspectors, short trains would have only 1 conductor and long ones would have 2 (one in each set).

    14. Caltrain is still running 4 car trains, in my experience (2 or 3 peak commutes per week SF to PA and back) fewer than half of the trains are back up to 6 cars. There are 15 or so idle gallery cars currently parked on storage tracks at 4th and King, only a few appear to be not ready for service.

    15. @Martin: The only advantage of a fleet of 4-car units would be (have been) to make use of a certain economy of scale. OTOH, the cost needed for stabling spare sets at the terminals would be significantly higher. So, overall, letting the long sts run comes cheaper at the end of the day.

      Overstaffing remains a problem, however.

    16. idiot doom spiral22 September, 2023 00:05

      > Caltrain reducing the length of their diesel trains was also super temporary based on my experience. Trains are mostly back up to 5-6 cars at this point.

      i saw a three car consist last night

    17. Without knowledge of details, we can only speculate on the possible savings. BART has 4x as many cars, but their savings are on the order $12 million / year in terms of cleaning and maintenance costs. Caltrain's fleet is about 1/4 the size, so possible savings would be around $3 million per year. However, BART's are due to ending the maintenance on a very old fleet. If BART had shortened trains using their new fleet, the savings wouldn't be nearly as great, as the new cars would in theory require less work.

      Caltrain's savings of running shorter trains composed of only the new cars are likely to be also much smaller than $3 million given the new state of the cars. Given the extra staffing required to move cars around, the savings just wouldn't be worth it the increased complexity. One could imagine that extra staff needs to be at end stations to move the empty sets out of the way and back into service.

      There actually is an example of where I saw this happen in the Bay Area. I used to see VTA stage extra light rail cars in Mountain View. Around 4pm, you'd see a 1-car train finish its run in Mountain View. The driver would then run the train to the staging track to pick up a 2nd-car with the shaku and roll back into the station. The subsequent run would be with a 2-car light rail train.

    18. Regardless of whether old or new (as with BART’s new FotF and Caltrain’s KISSes), in addition to wasted energy costs, every needless car-mile counts against their MTTBF rate as well as other mileage-based expenses (wear, inspection/service intervals, etc.). And unlike with OPTO BART, with Caltrain you’ve also got the expense (or savings) of an extra conductor on/off the clock for the entirety of every run (and then some).

    19. > And unlike with OPTO BART, with Caltrain you’ve also got the expense (or savings) of an extra conductor on/off the clock for the entirety of every run (and then some).

      Caltrain used to lock the doors and turn the lights off in some of the cars to shorten trains off-peak so that they could be operated with fewer staff. There’s nothing stopping them from doing this now (aside from perhaps concerns about labor relations).

  3. No mention of weekend service yet. Will it stay at 60 minute headways despite faster ridership recovery than weekdays?

  4. They're making a big deal about 75-minute SF-SJ locals, but this trip time was already achieved way back in 1978! Check on Train 170 on the linked timetable - and this was with 21 intermediate stops, compared to today's 20.

    1. I doubt many accommodations were provided to passengers needing assistance back in 1978. And ridership was likely much, much lower back then.

    2. I'm looking at SB train 122, a local that departs SF at 4:15 p.m. and arrives in SJ at 5:50 p.m. That looks like 95 minutes to me. Where do you see a local that's 75 minutes?

      Interestingly, the fastest "baby bullet" from that era departs SF at 5:14 p.m. to arrive in SJ at 6:23 p.m. in 69 minutes with 4 stops. Even our diesel trains from 2004 could do that in under an hour.

  5. To the "Implausible":

    • The trains do draw quite a bit of power. OTOH, according to the newest version of the data sheet, they have an enormous regenerative braking power, which means that this goes back into the grid. It is therefore the question how Caltrain and the power company deal with that.

    • Padding (for a local) should be a couple of minutes every 15 to 20 minutes. This approach keeps the schedules stable but still tight. These padded stations are usually major ones with (hopefully timed) connections. Typical in many places is one big lump of padding before arriving at the terminal station.

    • I do fully agree with level boarding. If there are pseudo-legal reasons preventing platforms at the level of the entrance doors, Caltrain should (yeah, it gets fishy) circumvent or blatantly ignore them (shoot first, and then ask…)

    1. Salt Lake City has level-boarding for their Bombardier Bi-Level cars, and I think that's same height as the lower level of the EMUs.

    2. Salt Lake City has "level boarding" in the same sense as the Northeast Corridor has "level boarding." If you're in a wheelchair, the platform gap is not compliant with ADA and you will need crew assistance to board or alight with the aid of a manually deployed bridge plate. Incidentally, the Bombardier floor height (635 mm) is slightly higher than the KISS EMUs (555 mm).

      What's needed on the peninsula is a fully compliant gap (2 inch horizontal by 5/8 inch vertical) to enable BART-like boarding without need for crew assistance. This is a hard engineering problem, not in the least because everyone thinks it's easy.

    3. Which part is difficult?
      - IIRC the KISS has a self-adjusting suspension, which should make vertical positioning highly consistent, with a precision comfortably inside 0.625" (~16mm).
      - KISSes come with gap filler plates under every door, right? (Where I live, FLIRT-2s and KISSes all have them.) These should be sufficient to cover any practically occurring horizontal platform gaps. (In the event American operators would for some reason require the fitting of ones longer than is customary in Europe, it is straightforward to do that.)

      Even if Caltrain and its contractors are limited to a medieval level of technology, the task is straightforward. Just hammer down a line of sticks where they want the platform edge to go, roll up a train, open the doors, mark on the sticks where the top of the gap-filler plate pushes up against them, then shovel filler material until its level reaches the marks on the sticks. The magic that makes this workable is already in the trains, unless for some unreason someone keeps mucking with the position (height and cant) of the tracks.

    4. Thanks for illustrating my point ("everyone thinks it's easy") with such gusto.

      There are no gap filler plates on these EMUs. There are steps that deploy down to 381 mm, yes, but these are not able to fill any sort of horizontal gap, quite besides being at the wrong height. A new mechanism that is compatible with both platform heights (203 mm and 550 mm platforms) would be needed for the multi-year duration of platform reconstruction. This device does not yet exist and may not be trivial to engineer and retrofit to this brand new fleet.

      Next, the relationship of platform to track is not static. Track is a flexible and dynamic animal and it shifts and settles over time by amounts that far exceed 5/8".

      Finally, even if you did have this dual-capable step/gap filler mechanism retrofitted to all the trains and if you rebuilt the platforms with adjustable (jackable) noses at every station, you would still have an issue of ADA compliance aboard the train where there would not be equal access to the bathroom. The bathroom might even have to be removed entirely to comply with accessibility regulations, barring the installation of interior lifts to travel step-free between cars-- a solution that was contemplated but ultimately abandoned because of technical and regulatory obstacles.

      This is a hard engineering problem, not in the least because everyone thinks it's easy!

    5. "A new mechanism that is compatible with both platform heights (203 mm and 550 mm platforms)"
      Absolutely not. You design your mechanism to work with your final design platform height [see note below]. If you don't have ADA level boarding now you don't need it during the platform raising period, and it is foolish to put R&D into a temporary solution. As Anon said the existing mechanism in use in Europe works fine.

      "Track is a flexible and dynamic animal and it shifts and settles over time by amounts that far exceed 5/8"."
      1) Then how does BART (and many other systems) maintain ADA level boarding now?
      2) Maintain your track.
      3) Switch to slab track at stations if needed.
      4) As Anon noted, the self-adjusting suspension makes this moot.

      "you would still have an issue of ADA compliance aboard the train where there would not be equal access to the bathroom"
      This is a separate issue from platform level boarding. If the issue is access to the bathroom then you only need a lift(s) in the ADA bathroom car. If there was a requirement for a wheelchair to reach the bathroom from a different car once boarded, then how will Caltrain be compliant when they put the trains they already bought in service?

      [Note]: Caltrain, and every system in the country, should be aiming for a platform height of 1219mm, compatible with the NEC. This would allow future orders to piggyback off of always much larger LIRR/MNR procurement. If CAHSR ever happens it will almost certainly use this platform height. If ADA bath access is an issue, then (again) it should only require lifts from the vestibules of the bathroom car. Looking at this sure makes the KISS seem like not a great deal space wise. Without bikes the Caltrain order would have 645 seats and 1 bath. An M9 A car would have 94 seats at 4 abreast, a B car with bath 88. Six A's and a B would be 652 seats and 1 bath, just edging out Caltrain's 6 car KISS's, and all of it step free on the train (including to the ADA bathroom!).

    6. "There are no gap filler plates on these EMUs."
      That is not an engineering problem but an idiocy problem.

      "A new mechanism that is compatible with both platform heights would be needed for the multi-year duration of platform reconstruction."
      1) Onux answered that correctly. The choice is between ADA compliance:
      - at some cost (platform raising) within a few years (with high schedule certainty, modulo known general incompetence);
      - at a higher cost (platforms still need to be raised) at some indefinite point in the future (developing the non-step steps).
      2) Incidentally, why the slow burn? I'd at least look in other parts of the tradeoff-space, even if that turns out not to be the right choice. (What affordances become available if the entire program fits in a double-digit quantity of days? Including e.g. on the lowest-traffic week of the year, a complete service outage, thus creating unrestricted 24/7+2 worksite availability.)

      "you would still have an issue of ADA compliance aboard the train [...] interior lifts"
      In the +550 part of Europe, KISSes have an accessible bathroom on the bottom floor (where the doors are) of a single carbody, and some wheelchair pictograms on the outside. Unlike regulators, wheelchair users are smart and thus they board into that carbody. The rest of the train being full of stairs is completely irrelevant. The problem is with the requirements, not with engineering.

      "The bathroom might even have to be removed entirely to comply with accessibility regulations"
      What is wrong with these people?

    7. Wow. Getting big big CBOSS energy here.

      "This is a hard engineering problem, not in the least because everyone thinks it's easy!"

      "Nobody's ever designed signals for both freight and passenger trains before. You might think it's easy, but it isn't!"

      How'd that work out, exactly? Who turned out to be right, again? Idiots at Caltrain, or European engineers who actually know how to go about building shit?

      $500 million later, Caltrain doesn't have normal working ETCS/ERTMS (which Stadler sells on thousands of vehicles) signal system.

      $2000 million later, Caltrain won't have normal working platform gap filling door mechanism (which Stadler sells on thousands of vehicles) for level boarding, and will not be able to operate reliable, predictable or cost effective service despite this "investment" in electrification, because the randomness, delay, and huge staffing overheads of failing to even consider ("plan for") level boarding.

      CBOSSing it like a boss!

    8. Part of me wants to throw in the Parisian solution to ADA access for RER/Subway which goes something like this: "If you're in a wheelchair, you can just take the bus"... (AKA... we're not gonna bother retrofitting/compromising trains for you). This wouldn't fly here, but it's worth noting that this comes from the country that brought us the TGV.

      The above joke aside, I'm curious how one handles high-level platforms on tracks where trains might skip and pass at 125mph. On one hand, 125 mph track is likely to be well aligned, but trains sometimes do sway back and forth. Risk is lower for level boarding at the low level that's closer to the track, but the sway is higher at the higher level.

    9. You can complain all you want about idiocy and "what is wrong with these people" and how you wish it were like Europe but it's just not. All of this is just refusal to accept that this is a hard problem.

      @Onux: you just can't have a train with a 550 mm platform interface dock at a 203 mm platform. To translate that to imperial units, that's a step over one foot tall, a lawsuit waiting to happen. This "just weld a plate over the step well" approach can't and won't work. You can build two entire sets of platforms (e.g. the Stuttgart U when they transitioned to high platforms, but unlikely here due to space constraints and astronomical costs) or you can shut the whole system down during construction, also unlikely here. You and others may not like it, but the path of least resistance is engineering the complex gizmo that handles both heights.

      Platforms where trains pass at high speeds are built some distance back from the train, leaving a large gap. This is how "level boarding" is done in the Northeast Corridor. You step across a large gap, while taking care not to trip or fall. To meet ADA criteria for boarding without crew assistance (which the NEC most assuredly does not!), you need a gap filler mechanism bridge the gap. See Florida's Brightline for example of how this works.

      That is how the KISS EMUs were initially engineered, but they were delivered with door plugs instead. If you retrofit the EMUs with high doors and gap fillers (these items are being delivered to warehouse storage), then they can dock at a high platform just fine. But that's where the trouble STARTS. During the transition period, trains will need to dock at both high and low platforms, which means ADA interior lifts, which are ridiculously over-engineered (800 lbs capacity) and impede interior circulation, quite aside from the annoying fact that current regulation prohibits their use while the train is in motion. This is the dirt road that Caltrain went down (to the point of all the EMU car bodies having structural attachments for these interior lifts!) before deciding to abandon the entire steaming pile of level boarding crap & proceeding to kick the can down the road.

      As to the ADA compliance of the current solution, it's grandfathered in, mini-highs and all. That's why the KISS EMUs are exactly identical in their platform interface and boarding process to the legacy Bombardier fleet. Zero difference, zero progress!

      I am convinced that the only feasible way forward is the dual height gizmo, preferably with platform edges that can be periodically jacked to the correct relationship with the rail. I described this four years ago here.

      It's easy to snipe at this solution, but if anyone has better ideas let's hear them, and without the snark and wishful thinking.

    10. "You can complain all you want about idiocy and [...] how you wish it were like Europe but it's just not. All of this is just refusal to accept that this is a hard problem."
      Please compare your own description of how Brightline solves level boarding, then. The arc here is not that the engineers of Europe (or Brightline) are more talented, or the laws of physics more cooperative (t)here. It's that the difficulties lie not in engineering a solution to the end users' (i.e. wheelchair users') problems -- solutions are in daily use in many places around the world -- but the difficulties are entirely the result of asinine requirements that help nobody. (I suppose Richard Mlynarik would add "except the consultants to bill for more money".)

      "what is wrong with these people"
      This is a perfect example, exactly because it has nothing to with engineering difficulty. Of the two scenarios on offer, namely:
      1) exactly 0% of passengers (actual or potential/latent) have access to a toilet (since there isn't one);
      2) 90+% of passengers, or even 100% of them if we can only trust that wheelchair users will board the right car, have access to a toilet;
      they deliberately prefer the first. It's not a tradeoff between user groups, it's not a tradeoff against some other consideration (e.g. the reasons subways don't have toilets) -- the first option is strictly dominated by the second, yet they choose the first. This is why I'm asking "what is wrong with these people".

      To harp on it some more: if you went up to some wheelchair users, and asked them "er, this toilet is not accessible to you, should it be rendered useless to able-bodied people as well, or should it stay in?", they would answer that it should stay. Because they are, well, normal people who just happen to be disabled. If the fairy goes up to them and says their cow died, they wouldn't ask for their neighbor's cow to die. Why should the regulations ostensibly representing their wishes answer in their stead "oh yes, my neighbor's cow should die"?

      Back on topic. Notice that so many of the oafish sub-requirements (causing engineering challenges) flow from the assumed requirement of having to serve stations with very different platform levels. Interior lifts to access the other door (or toilet), fold-down stairs vs. gap filler plates, etc. What happens if that requirement is flipped the other way, what sub-requirements flow from that? No mixed-platform-height service = at least the majority of stations will need to have their platform edges raised during a single whole-system service shutdown. The method described in the post four years ago (craning in precast platform-edge modules) is perfectly adequate for this. (And if the trains need any work done on them, e.g. replacing the fold-down stairs with gap filler plates, that can happen at the same time, when the trains are accessible since they aren't in service.) It's almost as if the difficulties lie not in engineering but in the requirements, in the assumption that "transition period" meant service to both platform edge heights rather than neither.

    11. "As to the ADA compliance of the current solution, it's grandfathered in, mini-highs and all."
      That's not how the ADA works. You can grandfather existing infrastructure, but existing infrastructure can't grandfather new projects. For instance, non-compliant ADA sidewalks does not mean you can build a new building without ADA bathrooms. If Caltrain will use the new trains without lifts, then the trains are ADA compliant without them, period, otherwise they couldn't have gotten federal funding or an activist lawsuit would already be underway. If the answer is that Caltrain will only deploy mini-highs for the car with the ADA bath, then that is exactly what you do with level boarding, you mark that car as accessible and then you don't need lifts in non-accessible cars. Don't try to say that all cars must be accessible, the country is full of buildings with signs at the front door giving directions to the ADA lift/ramp elsewhere. Since the EMU floorplans you posted at https://caltrain-hsr.blogspot.com/2016/09/emu-brochure.html show wheel chair spots in every car, but the bath in only one, and the lifts are not being installed, I have to say your argument that the ADA requires lifts to be fully untrue. If wheelchair passengers can board with assistance to a car that doesn't have a bathroom or a lift, then they can board via level boarding to a car that doesn't have a bathroom or a lift. ADA requires compliance via standards or reasonable accommodations, there is no get out of jail card that says if you are providing accommodations in one area you can ignore standards elsewhere (that would be like a building with an ADA ramp in the rear not providing ADA bathrooms inside; that is not how it works).

      I will admit I am not an expert in this area. If there is some legal maze whereby you can't provide level boarding because you can't move freely about the trains, then that's another reason the KISSs were a bad idea. Should have gone with FLIRTs (or Class 345s like the Elizabeth Line) - no ADA problem and you can get the same number of seats with a minimal increase in length. With open gangways you really solve the mobility issue, for everyone not just the disabled, and you gain standing passenger capacity for crush loads.

    12. Yes, FLIRTs have retracting steps; all of them running in Europe. The same applies for FLIRTs. The retractable steps expand as far as they need (sensors sense the platform edge).

      The air suspension does indeed change the height of the floor. However, it is used to cmpensate for different loads, making sure that the floor is always at the same height, no matter how empty or how full the car is.

    13. "you just can't have a train with a 550 mm platform interface dock at a 203 mm platform"
      I'm confused as to why this isn't possible. An automatic bridge plate that can drop down and act as a step is the solution to this problem. Looking at the mechanism on KISS delivered to Caltrain it looks like this is something it can already do--and even if it cannot, this is definitely something that the engineers at Stadler can figure out how to build.

    14. @Anon, the mechanism delivered by Stadler can't do that because it was not required to. It is a drop step at 381 mm that happens to retract. I'm sure they would figure it out if asked (along with their mechanism supplier, Bode) but they've not been asked most likely because Caltrain doesn't have an actual plan for level boarding and hasn't yet identified the need for this solution.

    15. Big. CBOSS. Energy.

  6. In their latest exercising of Options, Caltrain did order a prototype of a BEMU KISS, which would allow through running to the non-electrified segments. How well it will work is unclear, but I am relatively optimistic.

    1. I thought the BEMU would be shorter than the current 7-car trains. While you could run-through a short - say 3-car BEMU - up to SF, that train might fill up along the way.

      This might relegate the BEMU to an SJ-Gilroy shuttle service as proposed.

    2. The single 3-car BEMU won’t be here for years!

    3. The question is can a BEMU draw enough energy from Tamien to SJ and back to complete a journey from Tamien to Gilroy and back? That would seem to be a bigger power issue than fast acceleration, both for the line to deliver it and the BEMU to receive it. You can always shortcircuit the issue with long dwells to charge sitting at SJ, but that is a very inefficient use of expensive trains.

    4. Tamien to Gilroy is 29 miles, and there will definitely be charging infrastructure in Gilroy to ensure sets don't get stranded. 29 miles is actually a pretty short distance.

      What's interesting is whether Caltrain can run the Gilroy service of 4 trains using only 3 sets by deadheading the first run back to Gilroy while having enough capacity + recharging as you suggested between SJ and Tamien.

    5. For what it’s worth, Stadler claims its FLIRT Akku BEMU recharges in only 15 minutes. So time under catenary from Tamien north should be more than adequate with even a very short Diridon layover.

      Additionally, Caltrain’s BEMU announcement suggests they may also be installing a bit of recharging overhead in Gilroy (which may help ensure the BEMU can make it to Salinas & back):

      “… upgrades to the San Jose Central Maintenance Facility and Gilroy layover and station area to facilitate charging and maintenance.”

    6. The transportable charging unit Voltap by F+F can provide (with the FLIRT Akku) enough power within 10 minutes for about 50 km. So, setting up such a station (consisting of a transformer to the middle voltage grid (for galvanic separation), control system and a stretch of power rail; it is so simple because the battery charging system is on-board) at the terminal would allow to charge while the train changes direction.

      All the maintenance facility needs is tools etc. for the maintenance; charging can be done under wire (which is necessary for the straight units anyway).

    7. There's also the possibility of extending the electrification for say 2000 feet north of the Gilroy or south of Tamien stations to allow the trains to accelerate while still under the wire when most of the power is needed. Not sure if that's done in practice, but an extra million on some extra poles and wire could extend the lifetime of batteries, reduce electricity usage due to charging inefficiency, and save maintenance down the road.

    8. There's also the possibility of doing none of this extremely wasteful experimentation. Modern diesel hybrid buses have a lower carbon footprint than any other solution you might throw at the Gilroy branch, which generates about 1% of Caltrain's ridership. BEMU is a solution looking for a problem.

    9. Tongue in cheek: let's solve both Caltrain's funding problem and the Gilroy branch ridership problem together. Starting a bit south of Blossom Hill, the line runs through agricultural fields and other similarly "high-density" land uses, such as San Martin airport. Buy some of this land and build a town's worth of five-over-ones (i.e. in the low tens of thousands of units). Suddenly, electrification and double-tracking are justified (as well as funded).

  7. I wonder if a battery EMU could supplement catenary power with battery power to cut the peak load.

    1. No. The point of a BEMU is that it charges while under wire, so that the battery can then power the train when it is on an unelectrified line. If you drew power from a BEMU to meet the peak load of regular EMUs the battery wouldn’t be fully charged and it couldn’t compete it’s route.

      As Max noted above the power issue isn’t really an issue; regenerative braking reduces total energy use and the electric company sells energy not power: 10kW for 2 minutes costs the same as 5kW for 4 minutes. In any event electricity is a lot cheaper than diesel and I’m sure the revenue bump from attracting riders to fast service more than pays for the acceleration.

    2. Quoting from Clem’s original post, emphasis added:
      > the trains have to accelerate very aggressively, something the equipment is admittedly capable of, but ouch, that electricity bill! **Operating cost will depend not just on usage but also peak load, and high acceleration makes the peaks worse.**

    3. idiot doom spiral20 September, 2023 13:31

      > the electric company sells energy not power: 10kW for 2 minutes costs the same as 5kW for 4 minutes

      not at the scale of tens of megawatts. consumer electricity pricing obscures a complicated real-time energy market that balances instantaneous demand with supply from sources of wildly varying operating costs and limited capacity to rapidly adjust output. fortunately california has a huge battery bank and moss landing and the goons at CAISO are rather competent, unlike their PG&E counterparts

    4. @Onux: why don’t you explain to Caltrain exactly how “electricity is a lot cheaper than diesel”? Because the staff & board have been saying and fretting about the opposite! Have a look at Caltrain’s Strategic Financial Plan (SFP) Energy Procurement Strategy for Caltrain Electrification and see the gloomy news about how much more they expect electricity to cost.

    5. @Onux: at 104 trains/weekday, Caltrain's annual electricity usage will be about 1e8 kWh, at $0.21/kWh. That's $21 million bucks. With net metering, this could come down by about a third at most, but Caltrain does not have those agreements in place with PG&E as far as I know.

      Compare to $13.5 million spent on fuel in 2022.

    6. @Clem: does this number include the energy gained by regenerative braking?

    7. @Anon: The financial plan is worried because electricity cost went up 67% from 2017-2023; during that same period No 2 diesel in California went up 77%. The notes on the report state electric cost is for 128 TPD but the 2024 baseline doesn’t have 128 TPD diesel. The projection to 2033 shows electric prices rising every year, but diesel cost stays the same, as if diesel cost won’t go up for a decade straight.

      @Clem: How many trains per day was Caltrain running in 2022?

      If Caltrain will honestly spend more on electricity than it would for diesel for equivalent service, especially after net metering, then I don’t know what to say, California has screwed up more royally than I can fathom if electricity costs that much.

    8. Every so often you come across an article that say that installing new solar or wind is cheaper than continuing to operate coal plants - which seem to operate where electricity is already relatively cheap compared to California.

      As solar and batteries get cheaper, opportunities to utilize some of the RoW for power generation might be worth exploring. But realistically, the first opportunity for that might be to have a small solar farm in Gilroy to charge batteries during the day that will charge the BEMUs at night.

    9. @Max: "net metering" means the electricity meter is allowed to count backwards when regenerating into the grid. That would save about 1/3 of the total electricity bill.

      @Onux: Caltrain ran 104 trains/weekday in 2022.

    10. For its relatively much cheaper (eg $1.50/gal.) than on-road red-dye (off-road, untaxed) #2 diesel, Caltrain uses fuel hedging and/or futures contracts to help with price stability/predictability: Caltrain Fuel Cost Stabilization

  8. Having baby bullets survive as express is a GOOD thing. There are two reasons:
    1) Caltrain (and most rail services) need a headline timetable. What Caltrain has here is "SF to SJ under 60 minutes)". When Brightline opens the Florida route to Orlando, I'm sure the travel time will be one of the top bullets mentioned.
    2) Anytime someone looks into taking transit, the #1 item they check first is the travel time. They won't consider the train further if the travel time is significantly longer than driving. For a regional/long trip, I'm willing to adjust and optimize my time to reach my destination faster. More frequent but slower trains will merely serve to discourage new riders. In other words, I'd choose 2 tph vs 4 tph, if 2 tph travel time is 10 minutes faster.

    1. You may believe you would choose 2tph vs 4tph to save 10 minutes on a trip, but actual revealed preference is that you're either in a minority or that your belief in your preference would not accord with your actual revealed preference (don't feel bad about this, I hold tons of false beliefs, as does every human.)

      Re headline "SF to SJ under 60 minutes": nobody cares. Nobody! 1tph is meaningless on a regional service. (1tph SF-LA, not meaningless. 1tph Larkspur-San Rafael, meaningless regardless of speed.) You may think you care. You're wrong. You actually don't, because "you" (collectively) do choose frequency of service over headline average point-to-point speed pretty much every time in the sorts of service environments we're talking about here.

      (OK idiots who don't use the service and never will, like, for example, everybody at Caltrain and everybody on the PCJPB board, "care" about "SF to SJ under 60 minutes". That's because they're idiots, not because it would affect their entirely imaginary inclinations to ride the trains in any way that affects total ridership.)

    2. I won't disagree that I'm a minority, or that my experience with friends choosing to drive because Caltrain is just too slow is also a minority experience.

      Suppose local trains already run every 20 mins, and your option for a 4th train is either:
      a) Add baby bullet that saves 15 minutes
      b) Another local train that arrives, so all stations are served every 20 mins instead of 15 mins.
      A minority would choose A - according to you.

      However, suppose local trains already run every 20 mins, and your option is either:
      a) A baby bullet that saves 15 minutes
      b) Another local train that arrive every 15 mins
      Would the minority still be so small that we should choose A? What if we're going from 4tph to 5tph? Is the train every 12 mins vs 15 mins still better than an express? Is there ever a case where you'd support express trains vs having it be just all local like BART?

    3. "Is there ever a case where you'd support express trains vs having it be just all local like BART?"
      Dude, I've been at this for decades.

      I've travelled around the world with the specific idea in mind of finding and presenting examples of success for Caltrain to emulate. (Yeah, I was a fucking idiot for believing there would ever be anybody in anyway connected with Caltrain who would ever want to deliver "success", as opposed to delivering "earmarks" and "kickbacks". Fucking moron.)

      And yes, since you're busy ignorantly setting up strawmen, I do happen to think Caltrain should offer limited-stop service, lots of it, all day, every day ... at some now-pathetically-far-off point in the future. (Right now today on Planet Earth I'm finding it increasingly hard to argue with all-stops 20 minute headways, based on data, which is not a place I wanted or expected to find ourselves, but here we are.)

      Thanks for playing. Meanwhile, enjoy your once-per-hour, max-6-hours-a-day-only commuter railroading "Baby Bullet" experience! You have now had two decades years to enjoy it (because Caltrain hasn't substantially improved or revised its fucking stupid commuter railroading timetable once in over nineteen years (June 2004, to be precise). You're guaranteed to get decades more to enjoy it too, because that olde tyme commuter railroading shit service mindset isn't going anywhere. It's embedded well up past the upper arm in Caltrain's fundament.

      Highball on the green, sub-assistant conductor!

    4. Your proposal is unique and probably has some benefits, but it pretty much ignores the Santa Clara County market as the travel time is simply not competitive with driving even though it could be.

      If I look at your proposal in the context of my hypothetical question of the type of train to add when going from 4tph to 5tph, your answer is "Don't add a train". This sort of becomes a feedback loop where service is worse because fewer people ride the train.

      I'm also unclear where you get this statement because it's very inaccurate: "Caltrain hasn't substantially improved or revised its fucking stupid commuter railroading timetable once in over nineteen years (June 2004, to be precise). "

      In 2004, Caltrain's schedule was all local except for 10 baby bullets. By 2019, this was up to 22 baby bullets (and 4 on weekends), along with even more limited trains that didn't exist in 2004 (I don't have an old copy of the timetable, so don't know the exact number). Daily ridership in 2004 was about 24K vs 63K before the pandemic, so why isn't this "substantial" change in your definition?

    5. To Martin: I apologise, but I didn't understand the difference between the two pairs of proposals for what to do with the 4th train. (I understood both to start with 3 tph all-local, both options a.) are 3tph-local+1tph-limited, both options b.) are 4tph all-local? Surely I mistook something.)

    6. The debate is between how many trains should run local vs express. There are diminishing returns when adding more local trains because additional trainsets are needed and because there is too much capacity at the minor stops between major stations. Express service resolves this because shorter running times saves the railroad in operating costs and passengers in travel time. Traditionally, Caltrain runs a lot of express trains because its local trains (without electrification) are too slow and have low ridership, while BART runs local because it has no end of line stations with significantly higher ridership to justify skipping stops.

      With electrification, Caltrain is claiming an aggressive 75 min runtime end to end. If true, that's faster than all trains today except the baby bullet, which has the drawback of not serving a majority of stations. Thus it is viable to run local trains at 20 min headways, and no express trains are actually needed to serve capacity. From Caltrain's perspective, you need 30% fewer trainsets so you save money, while riders benefit by knowing that they only have to wait at most 20 mins for a train that will serve their stop on the corridor.

      The only issue is that Caltrain riders will protest against losing 4 tph, but in practice most OD trips are only served by 2 tph. If you have an express train from SJ or SF to Mtn View, Palo Alto, or RWC, you will never take the local train, so even there you effectively only have 3 tph. So when ridership recovers, 4 tph local should be better than 2 local + 2 express, since Caltrain only has 2 stations (Hayward Park and Bayshore) that are even worth skipping today, and those stops are on the shortlist to be disused.

      The summary is that instead of running additional trains that skip stops and underserve ridership, it's usually better to find ways to speed up the local trains. Increased acceleration from EMUs, level boarding to reduce stop duration, and curve straightening to reduce slowdowns are the primary ways to do so. If Caltrain ridership between SF and Santa Clara County recovers, then it's worth adding trains that skip most stops in San Mateo County while making all stops in SF and south of RWC. Then fo be cost efficient, SM county local trains can run between SF and RWC only, with timed transfers to the SC trains. These local trains can eventually run across Dumbarton to Fremont. All of this requires a 4 track RWC hub so it's not likely to happen in the near future.

    7. I guess my question was a general survey of what folks feel is the right express/local split at 3tph, 4tph, 5tph, etc...

      A more interesting way of looking at this would be to consider what travel time needs to be offered to attract riders who live ever further from the station. At one end of the scale, one would imagine that if you offered a 10 minute travel time between SJ and SF, you'd likely get riders from as far as 20 minutes away from both stations to take the quick trip. On the other end of the scale, a 90 minute trip time isn't going to get many takers who live 20 minutes away from the stations who would choose a 130 minute trip over driving.

      This takes me to my next point. If Caltrain wants to attract riders in a relatively low-density Santa Clara county where nearly everyone drives, a travel time that's competitive with driving is better than frequency. Those riders don't put as much value into whether a train arrives every 30 mins, 20 mins, or 15 mins because they are in control of their car (also, applies to companies which have shuttles timed to baby bullets). Unfortunately, the ridership today is too low for such service.

      However, once we begin filling the "75-minute" locals and "60-minute" baby bullets, perhaps Caltrain can investigate offering a "45-minute" express that - similarly to the 2004 baby bullet - makes only ~4 stops between SJ and SF. @Richard, how many stops could such a train make to cover SF to SJ in 45 mins with 79mph speeds?

    8. There will be no such thing as a 45-minute express as long as speed limits aren't increased. With no stops, an EMU is no faster than a diesel, and both cover SF-SJ in 47 minutes (with 10% pad) when making ZERO stops.

      There are significant gains to be made in the SF and SJ approaches, which are dreadfully slow and wasteful of valuable time.

    9. A 45 min express is also unnecessary. It is effectively impossible to drive SF-SJ in an hour, a quick check of a traffic app showed a drive time of 59 minutes at 11pm when traffic doesn’t exist. People would speed and make that shorter with no traffic of course, but that’s not an option most of the day. An hour trip (57 min start to stop) becomes the fastest travel option and fits in perfectly with a takt at frequency of 2/4/etc. tph. If you can get the travel time below this you maintain the hourly trip by adding stops, giving more people access to the express.
      “Run as fast as possible not as fast as necessary.”

    10. “Run as fast as possible not as fast as necessary.”

      Ugh. That was obviously supposed to be “Run as fast as necessary not as fast as possible.”

    11. @Onux. You're right that it's not possible to drive SF to SJ much less under an hour, but the point is that going to/from the stations will take extra time. The purpose for a 45 minute express would be to target passengers that could drive SF-SJ in say 70 mins, but still need 25 to get to/from the stations where the 45 mins express would result in a total travel time of 70 mins that's comparable to their 70 mins driving time.

      The extreme comparable example would be people would drive 90 mins from Gilroy to SFO to take a 50 minute flight to LAX that's still faster than their drive. If you then imagine that every 1 minute you shave from the baby bullet's SF-SJ travel time, you can provide a service that's comparable to driving for someone who lives 1 minute further from the station.

    12. @Martin,
      If to-station/train/from-station is 70 min, and driving is 70 min, then people will drive. Cars exist, and are better than transit for certain trip types, just as transit can be far better than cars for certain trips (such as travel to/from dense station areas along a linear corridor). Focus on the trips you can get. You can't chase ridership that won't materialize by squeezing minutes. Far more relevant to total trip time than minutes at the margin are things like wait time at the station (I say at the margin because a 60 min vs 85 min trip is definitely a big difference). Humans are not totally logical and 10 min waiting at the station has the same or more impact on reducing ridership as an extra 15 min travel time (many studies prove this). Proper takt has an enormous effect on reducing waiting time by allowing for knots (the bus arrives at 9:14 and the trains arrive at 9:15 and the bus leaves at 9:17 so everyone getting off the bus or train gets right on the other with no wait) and consistency of scheduling (its easy to remember departures at 9:10/:25/:40/:55, so if you trust the train to actually leave then you can get to the station right before).

      Caltrain's market is along the 101 corridor. A 60 min express from SJ to SF (and by extension 45 min from Mountain View, 40 min from Palo Alto, ...) would dominate other modes in travel time along the corridor. You don't need to be faster and there are operational advantages (Takt!) if you are not.

  9. Clem, if you can find and fire up your ancient numerical integrator for the Caltrain line, shoot me the inter-station intervals for KISS-6MW-79MPH (we have KISS-4MW-79MPH and KISS-6MW-100MPH but not that one) and I'll add it to the Taktulator (first time I'll have touched that terrifying junk PHP code since 2012) I bet we'll get an excellent match between Caltrain's sadly-unachievable-in-practice promised trip times and that model of train performance with 30s dwells (unachievable, sadly) and 10% pad (better than they've ever managed to date, pathetically.)

    Modelling the locals as KISS-6MW-100MPH (the 160kmh top speed being almost irrelevant for our close station stops) and the limiteds as KISS-4MW-79MPH (so as not to overspeed north of Redwood City) with 30s dwells and 10% pad is already pretty much on the nose.

    1. BTW tweaking to 2tph local 2tph "Silicon Valley Limited" and 0tph "hello 2004 called and wants its timetable back" "Baby Bullet", I can get down to a fleet of 13 trains and up to a score of 120.

    2. [Didn't intend to post as "Anonymous", though it's always obvious when I accidentally do so, isn't it?]

      I meant 13 trains and score of 121, by virtue of uniform even 15-minute headways between Redwood City and San Jose.

      But wait! A reminder that 3tph all-local service (SAME AS BART!) scores 121, with a fleet of only 10 trains (80% utilization), assuming the same (unrealistic) 30s dwells and (optimistic) 10% pad and 6MW acceleration as Caltrain's iffy 75 minute SF-SJ local promise. HEADWAY IS KING! Repeat: HEADWAY IS KING! Run it all day, every hour, every day. Just do it. Don't overthink it! (FYI going from 2tph to 4tph all-local 6MW 30s 10% only ups the score to 125, at the cost of 3 more trains.)

    3. Richard, I do intend to provide these numbers to you after I dust off the old performance calculator. Stay tuned. Also, weird: when I click on your link above, I get a score of 119. Was 121 a typo?

      When we talk about 4MW and 6MW, the run times were calculated for a six-car train, per the original Caltrain order. The seven-car train has a rated power of 8MW, which is even sportier than the 6MW version (as it now has 8/7 MW/car). It's a friggin' rocket with an adhesion factor likely north of 64%.

      The benefit of "pedal to the metal" is worth about 4 or 5 timetable points. Remains to be seen if it is worth the extra cost of electricity, which as discussed elsewhere is not cheap.

    4. Update: I found the old code and it still runs...

      Back in 2011, I modeled the 6MW train with 6 cars and 42% of mass on driven axles, roughly based on the Zurich S-Bahn KISS, which had just entered service. Caltrain's KISS has only one more car, but has 16 powered axles (to Zurich's 8) and double the power rating! So even my 6MW train may be conservatively simulated because the real thing has even more power and higher adhesion that allows it to lay down all that power to the rail at even lower speeds. That 8MW figure is a continuous rating, and the short-term rating might be well north of 10MW.

      There is finally an EMU swifter than the legendary Jersey Arrows!

  10. On Slide #5, it reads:
    "Electric trains (19 trainsets by end of 2024; 23 trainsets by 2027)"

    Have there been more details on why the final 4 trainsets will take another 24 months for delivery? Are those part of the expansion plan to DTX, or do they reflect the BEMUs?

    1. Note: only one BEMU (singular) has been ordered … and to my knowledge nobody has yet made public a schedule for its production, testing & delivery. The extra four 7-car trainsets were ordered along with it as part of an expiring option to the original 19 EMU trainset order.

    2. The Stadler factory which does the carbodies is quite busy, so, the production of the carbodies has to be figured into the current production schedule. That requires its time.

  11. Not sure if anyone has reported on this yet: https://www.trains.com/trn/news-reviews/news-wire/caltrain-completes-installation-of-poles-for-electrification-project/

  12. I honestly think Caltrain should follow BART and just run 3 tph all day. This makes the schedule incredibly easy to remember and facilitates Millbrae timed transfers. (As a side note, Antioch trains need to run to Millbrae in place of Richmond trains. Intuitively the full time line needs to serve Millbrae and the added tph on weekdays would help a lot.) As ridership recovers, then add the limited stop trains in between the locals. Eventually, you get 6 tph at the SF stops and all stops south of RWC, where congestion is heaviest.

    1. Seconding the need for Antioch trains to Millbrae, for the 6 tph weekday service as well as the yellow to orange line timed transfers. I doubt though that Caltrain would prioritize Millbrae transfers even if they were on a 3 tph schedule, seeing as their weekend schedule now is 1 tph and transfer times still suck, especially southbound.

    2. Just a reminder, all red line trains (meaning all trains) to/from Millbrae now stop at SFO, which adds another 12 minutes to the already slow red line trip times. Airport passengers are clearly more important than commuters.

    3. "Airport passengers are clearly more important than commuters"
      Yes. That is, in fact, clear. There are 80% more of them.

      The good news is that nobody has to speculate! There are data. Using these data, you can accurately target your ire and your everlasting burning hatred and your criminal prosecutorial resources on the individuals responsible for the BART extension to Millbrae! (Your search terms include "Steve Heminger", "Takis Salpeas", "Quentin Kopp", "WSP" (aka "PBQD"), and "Tutor-Saliba". But start with "Steve Heminger" — anything local that involves over a billion dollars of outright unambigous fraud in anything remotely connected to transporation inevitably involves chief enabler "Steve Heminger".)

  13. The key thing to remember when it comes to BART and Caltrain schedules is the issue of representation. The reason BART gives two fingers to Caltrain riders at Millbrae is that SMC and SCC have no representation on the BART board and most BART staff live in the East Bay so they couldn't care less. In their eyes, the SMC extension exists purely to get SF and East Bay riders to/from SFO.

    Similarly, the reason that south SJ and Gilroy trips get such attention and budget is that almost half the SCC population is served by these station. Even though the ridership potential is low, the voters there are paying for a large fraction of the Caltrain budget through measure RR and VTA sales taxes. Their representatives (if they want to get re-elected) have to fight for more service whether it makes sense or not. Think of Gilroy for Caltrain as the equivalent of Livermore or Antioch for BART. Of course, it makes more sense to serve this market by bus (and the existing VTA route 568 is actually pretty good, with TSP along Monterey Highway), but for whatever reason, South SCC voters really like the idea of Caltrain Rail service, even if most will never use it, so that's what they get.

    1. I would draw a very strong distinction between southern parts of San Jose and the rest of the Gilroy branch. Most of those riders and taxpayers and voters live between the Cahill St station and Blossom Hill, and Caltrain service should very much be extended and electrified in that corridor, as discussed in my San Jose post.

      The connection at Millbrae is bad because Caltrain runs an irregular timetable, and will remain bad because BART runs a different base takt (20 minutes instead of 30 for Caltrain). No need to invoke political representation when simple math will do.

    2. Electrifying to Blossom Hill is not going to happen unless HSR gets its act together someday. In the meantime, we need to accept that diesel/BEMU service on the Gilroy branch is here to stay. Extending it up the Coast line terminating at Great America and/or extending Capitol Corridor trains to Gilroy would be more achievable improvements in the medium term.

      As was suggested in other comments to this post, the best solution to for the connection at Millbrae would be for all yellow line trains to extend to Millbrae. Since BART has decided that all Millbrae trains should run through SFO anyway "to avoid confusion to SFO passengers" at least in this case the 7 mins extra trip time would be compensated by the increased 6+tph frequency. As I say, it is all about priorities.

    3. Perhaps we're overthinking the Millbrae connection right now? Both BART and Caltrain have temporarily shortened their schedules to survive the ridership collapse, so aligning their takt is down on their priority list.

      Clem's post got me thinking though. Do you think there should be an effort for some board members to serve on both Caltrain and BART boards? This could potentially help with accountability of scheduling and future 2nd Transbay Tube (if that ever happens). Then again, it could also be a way for one side or the other to infiltrate their "enemy". ;)

    4. "The connection at Millbrae is bad because Caltrain runs an irregular timetable, and will remain bad because BART runs a different base takt (20 minutes instead of 30 for Caltrain). No need to invoke political representation when simple math will do."

      There I fixed that for you. You're welcome. Caltrain 3tph 6MW 30s dwell 10% pad Millbrae-centric 10 trains! TEN! 119 Taktulator score (re Clem reply to an earlier comment, 121 was a cut-and-paste typo.) Caltrain service optimized around the most important thing on the planet, the 3tph BART Richmond—SFO—Millbrae service for all ~10 (TEN!) transferring suckers a day who get sucked onto the billion dollar testament to corruption that is the Millbrae BART extension.

      (Fun fact: daily BART entries and exits at Millbrae were "predicted" by Steve Heminger's very very very special friends at PBQD, who are also the corrupt WSP-rebranded scum who destroyed any hope for California HSR, to be over 30000 entries+exits per weekday. Actual pre-pandemic Millbrae BART ridership? A hair over 7000. What's a 75% ridership shortfall between friends, huh? Last month? 4403 yeah sure, you want to operate 9tph to serve that ... NOT! 30000 "prediction" was total complete disgusting sleaze mendacity, in service of nothing but private profit, everybody knew it at the time, and still they get away with it over and over and over while the planet dies.)

      PS That timetable can't actually be operated because of the single Caltrain-controlled track bottleneck between SJ Cahill Street and SJ Tamien. Most Taktulator submissions we've ever seen here have that same flaw. I choose to ignore it 99% of the time.

      Cutting SJ—Tamien service back to 1tph works with the same 10 train fleet, but lower utilization (76% not very good 80%) and marginally lower score (116 not 119) Tamien is a tail wagging the dog, as is always the case with everything at Caltrain.

      But 10 trains. TEN! There's really no reason not to Just Do It.

    5. idiot doom spiral22 September, 2023 17:32

      > the 3tph BART Richmond—SFO—Millbrae service for all ~10 (TEN!) transferring suckers

      is that a typo? i find it hard to believe that I am on order of 1% of total milbrae transfer ridership.

      i expect that a synchronized takt would greatly increase ridership, as their really is no other way for car-less suckers like me to get the east bay/the mission from the penninsula. especially on the weekend when the DB is not running.

      OT but running the DB on the weekends should be amongst the highest priority short-term projects in the bay. on the weekend the fastest way to get from palo alto to union city is a miserable 90m bike ride. maybe link21 can divert some of their "study" dollars into actually connecting the bay.

    6. The irony is that, even if Caltrain switched to Richard's 3tph all-local service with timed transfers to BART at Millbrae, it would still be almost as fast to get off at San Bruno and walk/bike the 0.5 miles to San Bruno BART since it takes BART 7 minutes longer to get from Millbrae to San Bruno (via SFO) than Caltrain.

      With BARTs current attitude it might be time to give up on Millbrae transfers completely and have Caltrain express trains stop at San Bruno instead.

    7. "the 3tph BART Richmond—SFO—Millbrae service for all ~10 (TEN!) transferring suckers
      is that a typo? i find it hard to believe that I am on order of 1% of total milbrae transfer ridership.
      Sarcastic hyperbole. The "~" in "~10" was the super subtle tell. That, and the spittle-flecked ranting that is associated with anything connected to the BART extensions.

      Whatever ... it is it's not the 10k a day or whatever (I can't recall exactly and I don't want to make myself sadder by digging out my physical BART SFOX FEIR to precisely document the lie) Caltrain/BART transfers "predicted" in order to goose the US Federal Transit Administration of more than three quarters of a billion dollars in sweet sweet sweet earmarked cash. Direct convenient frictionless transfer from everybody's federal taxes right into the deep private pockets of HNTB, PBQD, Tutor-Saliba and friends, no detours, no switchbacks, no delays, fully coordinated and perfectly executed.

      It's nothing that anybody sane would plan anything around, let alone consider. Negligible. And yet here we are.

    8. "With BARTs current attitude ..."

      Dude, you really need to get some perspective. It's not about you.
      You're like, one dude, one who likes that one thing (Millbrae transfers) that nearly everybody else has determined is utter shit.

      Anyway, "BART" — the scheduling, the operations, the train control, those sort of useful people — have a pretty fuckingly amazingly good "attitude", and pretty much always have in the decades I've lived in the Bay Area Rapid Transit District. They're dealt a shit hand by the assholes who "design" and build BART extensions, they're fucked over night and day by the priorities of the idiots representing Orinda and/or (why not both!) Parsons Brinkerhoff on the BARTD Board, but, day to day, they deliver pretty good (if painfully and unnecessarily expensive) service that has always done a fairly good job of delivering as much useful service as possible under stupid constraints, and most recently has been really doing an excellent job of looking at the data and proposing and implementing useful service improvements based upon actual data, ie upon what non-insignificant numbers of actual human beings here right here right now on this planet might find useful to them.

      But sure. "BART attitude". Whatevs. I'll take that a million times over "Baby Bullet 1tph 4eva send more money" Caltrain Attitude.

    9. I will say that I agree with you that BART scheduling has generally done an amazing job dealing with the extension designs and have made a great improvement to service with their new service pattern. The new pattern certainly increases service where it is needed (yellow line) and makes cuts elsewhere to make sure ridership demand and frequency are balanced.

      However, I still believe increased service to Millbrae is justified in terms of a ridership versus frequency perspective. I don't think anyone is arguing for 9 tph at Millbrae from having both red and yellow lines, but the 6 tph from switching the yellow and red lines would give Millbrae service on par with other stations of equal ridership. While you may see 4403 entries+exits as negligible, especially in comparison to the ludicrous numbers given when the extension was proposed, split over 3 trains an hour this gives Millbrae an "entries+exits per day, per train per hour" ratio of 1468. Looking at the ratio of entries+exits last month versus the tph in the new schedule, Millbrae has one of the highest, only surpassed by Embarcadero, Montgomery, Powell, Berkeley, and Dublin. Millbrae has almost double the average ratio of 781, which means doubling the frequency to Millbrae would put it on par with the frequency of other stations of its size.

      Other stations with high ridership relative to demand are either some of the highest ridership stations in the system, most of these stations having frequency at almost full capacity, or are Dublin station, which is at the end of a long spur with very little ridership up until Dublin. Millbrae seems the most easily fixable of these ridership/frequency imbalances, with a simple switch of line color for one stop between SFO and Millbrae evening out frequency to match ridership.

      Maybe there is something I am missing. The folks designing the BART schedule certainly aren’t stupid, and I’m sure a yellow line to Millbrae is something they considered. Maybe the way trains are stored and turned around different at SFO and Millbrae, and therefore Millbrae is prohibitive to 6 tph frequency. Maybe extending the yellow line that far would require an extra train which makes the extension not worth it. In any case, barring these concerns, it still seems to me that Millbrae is more deserving of frequency, and my playing with the data in excel seems to confirm this as far as I can tell.

    10. "Nash", I (obviously, because I try very hard not to express or even hold any opinion sans data) am quite aware of BART's station origin/destination data, present and historical.

      My take is that you and "jpk122s" and fellow Milbrae-appears-on-a-map-as-a-TRANSIT-CONNECTION-therefore-IMPORTANT people need to take your case about BART service to Millbrae up with people at BART, not here. At present, the BART people at BART — possibly guided by origin/destination data, perhaps guided by revenue hour costs, perhaps guided by train or crew allocations, perhaps guided by fare revenue data, perhaps forced to do insane things by political directive, perhaps doing insane things because the words "airport connection" seems to make upper middle class twits completely lose any rationality they might have had — seem to have decided that BART service south of Daly Ciy should all go via SFO (recall: 180% the riders of Millbrae), and that running trains further on to Millbrae is something that BART can justify in terms of BART trains and BART crews and BART costs and BART revenues only if SFO service isn't short-changed by doing so.

      Now it is true that Millbrae, along with Dublin and SFO, are the only BART extension stations whose ridership is not utterly, abjectly, catastrophically bad, with Millbrae accounting for 1.4% of BART's present riders. That's bad and utterly inexcusable and criminal (recall: 30k/day used to "justify" its construction vs 7k/day actual pre-pandemic!), but it's not as catastrophically bad as the other extension stations Bay Point, Antioch, Colma, San Bruno, Berryessa, South SF, Warm Springs, West Dublin, Milpitas, Castro Valley, Oakland Airport, North Concord, Pittsburg Center, which are all total fucking sub-1% turkeys.

      But the thing is that Millbrae doesn't lie on the way to anything else, unlike, say, San Bruno (abject failure), Warm Springs (abject failure), Castro Valley (abject failure), and so you don't get service there "for free" on its way to some other "successful" ("success" grading on the curve of abject BART extension failure and fraud, comparative "success") destinations such as SFO or Dublin. You get service there by actively cannibalizing frequency to SFO, and, recall, the SFO station had 1.8x Millbrae ridership both pre-pandemic and today.

      I personally believe that BART running 9tph (vs 3tph) south of Daly City is an insane waste of resources and is unjustifiable by any ridership or revenue data, but is forced by the upper middle class fuckwittery associated with "airport rail". If I thought it would do any good, I'd take that up with BART. Likewise, if you passionately believe that 6tph or 9tph or 16tph should serve the bustling BART/Caltrain Pan-Galactic Inter-Modal Nexus in the Millbrae Segment of the Galaxy, I suggest you take it up with BART.

      You could also take it up with Steve Heminger who, after a career of inflicting hundreds of billions of dollars of harmful freeway and BART projects on the Bay Area at MTC (including the Millbrae BART disaster), now continues his legacy of poisoning the planet while comfortably sitting on the Caltrain board.

    11. Fair enough. I thought it was relevant to discuss how BART is altering service to Millbrae, given that the press release that is the focus of this post mentions BART’s service change, seemingly in a negative light. I only ask because I am curious to see if any of the smart commenters on here can give me a reason for BART’s decision. I think you’ve probably answered my question by saying that SFO frequency cannibalization is the most likely answer. I might see if I can get an answer from some BART related forum, but I doubt they’ll have a more satisfactory answer.

    12. Here's the list of BART branches by Aug 23 weekday and Sunday entries and exits:

      * Richmond - Ashby: 32k / 16k
      * Antioch - Rockridge: 35k / 13k
      * Dublin / Pleasanton - Castro Valley: 11k / 5k
      * Berryessa - Hayward: 22k / 10k
      * SFO / Millbrae - Colma: 19k / 12k

      So BART's current service (6 tph on all branches except Dublin) is in line with the data. Richmond trains do need to be longer, as that branch has always been the busiest in trips per mile.

      SFO itself is one of BART'S non-commute focused destinations and has recovered ridership faster. It's now a top 10 station on weekdays and top 5 on weekends, and BART's mandate is to target areas with potential growth. Hence the increase in effective tph by having all trains serve it before Millbrae.

      The area around Millbrae itself has developed a lot, with shops and TOD the past decade, and its location next to El Camino makes it a reasonably effective Caltrain / Samtrans hub. So there is an argument to increase service to 6 tph by running both the Richmond and Antioch lines for 6 daytime tph (the extra Pittsburg / Bay Point trains still terminate at SFO). However that requires an additional trainset from the Concord yard, plus probably reopening the island platform at Millbrae, which isn't cost neutral. Nonetheless, I do think Millbrae needs to be served by trains to Antioch full time instead of the Richmond line, which has a cross-platform transfer in Oakland.

    13. The area around Millbrae itself has developed a lot, with shops and TOD the past decade

      Just remember that in reality it is transit-adjacent development, and living within an easy walk of BART (i.e., within real walking distance to matter to most people) in no way "induces" [sic] BART patronage. (BART's reaching downtown San Francisco itself does support and attract patronage.) If enough new residents by Millbrae work in downtown San Francisco ...

    14. South County (GIlroy and Morgan Hill) did not have scheduled transit service before the creation of the Santa Clara County Transit District, which later became VTA. So this is why Caltrain service is so sensitive to their elected leaders and why they have lobbied so hard to get that 4th train. Cutting South County service might save money but it would be politically unpalatable. Note that the VTA 168 express bus was replaced with the limited-stop 568 which runs all day. Caltrain is doing the right thing by terminating those trains at San Jose Diridon with a timed transfer. It should also apply for a waiver to hang on to the diesel locomotives after 2030.

    15. Are they going to have four crews drive the quoted 43min trip from Gilroy up to San José and park it for the day (at CEMOF)? The current slideshow does suggest that:

      > 4 daily roundtrips (schedules tailored to survey feedback from South County residents in June 2023

      The other option seems to be to dead-head trains back south to Gilroy. There's currently not a large enough service gap for a train to make the trip one way, turn around, and come back - the first northbound arrives at Diridon at 6:42a, and the last leaves Gilroy at 7:29a.

    16. Gilroy is about 28 miles away from Diridon, so while there's only 47 minutes to turn the train and run it back to Gilroy, a non-stop deadhead or an express with passengers, could make that trip in about 35 mins. Caltrain could also alter the schedule for the 1st and 4th train to space the trains out and add an extra 10-20 mins which would increase the margin.

      Either way, three trainsets and three crews could reasonably implement a 4 train Gilroy schedule that's not much different from today.

  14. The approximation sign, or "tilde," used before 10 means the expression actually is inexact, but it also conveys reasonable closeness as well as being an inexact value -- exaggerated here, 8-12 (50% MORE! WOW!).

    Meanwhile, actual size or importance (for PR?) is enough to merit some published material on this subject, with a schedule to help those who transfer at Millbrae.

  15. Shock news! VTA’s 6-mi. SJ BART extension cost rises to over $2b/mi., opening delayed till 2036: San Jose BART extension will be further delayed and cost more

    1. If the cost starts to get worrisome, we can always raise bridge tolls again.

      Don't overlook another, "gentle," sales tax increase, too, under 1% to begin.

    2. Past history suggests MTC will pay for this by raiding any and all funds that would other have gone to Caltrain and other transit service.

    3. Latest news says Bay Area residents are now less likely to support new taxes or tolls for transit, particularly BART/Caltrain-like “commuter rail” systems: Bay Area Residents Are Pulling Their Support for Public Transit

      VTA’s downtown SJ BART dream has always been underfunded since its conception in the late 1990s, causing it to be split into phases (Warm Springs, Berryessa, and now, stupidly, Santa Clara Caltrain via downtown SJ). It’s not going to be easy to sell yet another SCCo. BART tax (this one will be the last, we promise!) or for MTC to find another $3 billion hiding in the couch cushions and/or to shift/borrow/steal from other projects!

    4. Drunk Engineer, while MTC indeed does this and has put BART up front, it has with Caltrain as well, aligning with the contractor enrichment theory of partially explaining MTC operations. Among the biggest uses of all among the bridge toll increase spending is not just the biggest, new BART cars, nor the second biggest, BART to San Jose Phase 2, but the third biggest, Caltrain Downtown Extension. But wherever it can find the money, it's likely BART will get more from MTC, "to help get VTA and BART to the end."

      Who else is grasping, even more, is VTA. That includes when VTA was allocated a fraction of funds from sales taxes for thirty years. VTA sought to get all the funds it could each of the first ten years. Anyone functional can guess what VTA would try to do later after getting all its share up front. Grasp, Grasp, Grasp, More, More, More.



      and so on



    5. @Reality Check, don't overlook that like some other craven or warped types, greedy MTC wants to toll more freeway lanes in the area that get substantial use including by commuters, as well as what has long been anticipated, start tolling arterials, too, beginning with tolling arterials parallel to tolled freeways to catch those who choose to drive parallel to avoid paying tolls on the freeways. Of course it was coming, "ideally" throughout the Bay Area and Sacramento greater metro area and overlapping exurban commuter sheds. The now-progressing Corridor Express Lanes aren't enough motorist plundering, any more than current bridge tolls. Will VTA want these sooner, or MTC for other favored uses (not limited to transportation, its legitimate purpose), that is another relevant question. If it happens, expect VTA's reach.



  16. I requested Caltrain's ridership statistics for the South County stations and any estimates staff had made for the fourth run. They didn't provide any estimate for the fourth run (who knows if they even bothered trying to estimate it), but they did provide ridership estimates for stations south of Tamien for every train from 1 April '22 to 31 August 2023. It looks like they're getting around 200pax/d on the section with three-train service. Here's the document they provided, if anyone is interested:


    1. @Evan: weird that Caltrain apparently didn’t bother filtering out all the non-Gilroy train ridership data from the Gilroy train ridership report you requested.

      I don’t know if you’ve already seen it, but staff published results of their survey of Gilroy riders’ priorities along with average ridership by train & station in August. See pages 10 & 11 of this slide deck.

    2. It's not "weird", it's an archetypical "hostile dick move" by an agency that hates the public and hates openness.

      I mean, a 2.9mb fucking *PDF* file from which one is expected to extract 206890 lines of station/train-number/date/passengers data of which only 5235 are non-zero?

      "But the plans were on display…"
      "On display? I eventually had to go down to the cellar to find them."
      "That's the display department."
      "With a flashlight."
      "Ah, well, the lights had probably gone."
      "So had the stairs."
      "But look, you found the notice, didn’t you?"
      "Yes," said Arthur, "yes I did. It was on display in the bottom of a locked filing cabinet stuck in a disused lavatory with a sign on the door saying "Beware of the Leopard"."

    3. The best part was that I requested any projection for the fourth train in the same request, and they neither said they didn't have any nor gave me one. Did they miss that part of the request in the bustle to take a few weeks to get me these numbers, did they pretend to miss it, or did they not have any? :)

      My thought here was that if we're lighting money on fire to run a fourth train, we might as well at least run a reverse-commute train when the shuttle service starts in 2024. Timing-wise it'd work out for the first train to also be the last train, and run Gilroy-SJ-Gilroy-SJ. My curiosity is satisfied with this and the deck Reality Check posted; I'm convinced ridership wouldn't be that much worse than existing commute service.

    4. To Evan - I think that would make way more sense. If they were a bit smarter, I could also see them run a simple all-day shuttle service between Gilroy & San Jose depending on the freight traffic with only 2 shorter trains and minimal staffing.

  17. Oops!

    San Jose pays $24M for site OK’d for homes to enable HSR
    “Critical” site needed for station track throat trades at 58% premium

    San Jose and the Valley Transportation Authority (VTA) will pay a steep price for the site of an approved apartment tower that could have blocked a high-speed rail line into Diridon Station.

    The city bought two parcels that make up the 1.12-acre property at 32 and 60 Stockton Avenue for $23.8 million — or 58% more than they traded for last year and in 2021. VTA will assume the purchase and sale agreement from the city.

    The seller was the Thang Do and Chunhong Liu Revocable Trust, the estate of the late architect Thang Do.

    The deal puts an end to plans by Do and San Jose-based Urban Catalyst to build what they dubbed Apollo — a 20-story, 471-unit apartment tower with 7,600 square feet of shops and restaurants that would replace an auto body shop and car wash at a cost of $100 million.

    The authority seeks to put a 61-space parking lot for Amtrak and Capitol Corridor rail service riders on the site, which would also be used to widen the railroad tracks.

    It’s unclear why the city or VTA didn’t step in to purchase the properties sooner at a lower cost. Instead Do bought them, and the city approved his and Urban Catalysts plans in November.


    1. While the big Diridon rebuild doesn’t have an approved EIR, the HSRA has had an approved SF-SJ FEIR/S since August of last year: Nation’s First High-Speed Rail Project Now Environmentally Cleared Between San Francisco and Northern Los Angeles County

    2. (repost due to bad HTML tags as I really need a preview button)

      The CAHSR selected alignment (Alternative 4) and station design do not really impact the lot in question as they keeps the station footprint and platforms almost the same as today, just modifying some trackwork and adding pedestrian overpass for circulation. They do show this lot used for an AMTRAK parking lot, but that could be relocated elsewhere. See Station plans PDF (p26)

      This lot is really needed for the Diridon Integrated Station Concept (DISC) plan which is a separate idea led by SJ/VTA to elevate the station. They are in the process of deciding whether that idea is feasible. It would require relocating CEMOF and the PG+E substation, as well as elevating all the UP and Caltrain tracks so it would be crazy expensive for not much benefit over the simple CAHSR plan.

    3. It is interesting that the high-speed rail program, for something that's a peer of the BART extension locally if not often viewed as worse, at least pays attention to parking needs while BART and other parties want to build housing, straying from its mission but which is done elsewhere along with commercial operations, and should have been done on top of the new Transbay Terminal, also hosting BART, MTC, Muni, and so on. Parking is needed in the real world, not just new housing to exploit a current fad pressing for more housing and to secure new sources of middle- and long-term money.

      The New Diridon "Grand Central Station of the Western USA," as with the Google village, is viewed cynically by the grown-ups and will continue to be until the plans are seen, then progress is made. We should expect easier and faster progress than with the local colossal HSR project equivalent, the VTA BART extension to San Jose and Diridon Station, and particularly since it has been attempted before with other funds, station proponents should be wary of VTA trying to divert or redirect (grab) station funds for the BART extension, too.

      If there is a Change of Plans with the new station later, the lot could always be flipped or become housing, after all.

  18. What's surprising is that the city and VTA sought it in part for parking instead of the high-speed rail project, which has sought parking, but has been in conflict locally about development on station-adjacent lands, as at Millbrae.

  19. https://sanjose.legistar.com/View.ashx?M=F&ID=12376781&GUID=65DA7587-3D74-4260-8C24-F5B9859C537F

    2 Caltrain tracks + 4 HSR + 3 non-electrified ??

    1. Thanks for posting that. I know HSR's designs in the approved EIR/S didn't take that property. This is for San Jose's conceptual station. HSR plans to swing wide into the arena's parking lot to ease the speed restrictive curves around the CEMOF. The city's conceptual plan takes out the approved apartment because the city is scared to death of the Sharks, who demand nothing but surface parking. Those acres of surface parking are sacred land in San Jose. Priorities, you know.

    2. Sorry for the messy posting. It's Monday morning. Correction to my previous posting- The HSR plan doesn't take out the BUILT project next to the approved project.

    3. As far as I know, even though it came from the city, the picture @Drunk Engineer showed does not reflect the latest proposal for DISC. That is described in the following (long) PDF See p.71 for the property impacts showing that the existing apartment building (Vespaio) will probably remain.

      @Michael note that the parking lots east of the tracks belong to Google now. While the Sharks are demanding SJ keep the amount of event parking available that was guaranteed when they signed their lease, the goal of preserving the Google Office project is what is keeping the plan mostly within the existing ROW north of the station.

    4. San Jose has a lot of practice building "high-rises" over many floors of parking. Buildings in the (don't hold your breath) Googleburg could be built up, around, and over the railway. More building floorspace, no trains to see, train alignment allows decent speeds in and out of station. That's called innovation, and should be embraced in Silicon Valley's Capitol of Innovation. It also does no evil. Win win!

    5. Michael, if you're including building atop the new train station or Grand Central of the Western USA or the galaxy, etc. as in Miami and as should have been done in San Francisco, we'll believe it when we see it. Note the Miami Central station has elevated tracks and elevated tracks is what a number of people in the South Bay want to enable the best circulation across the rail right-of-way in the station area. More bragging rights for those wanting them and small-timer salve in the South Bay, I guess that's an objective, too. The actual Googleburg around the station area always was nothing but words until built and occupied.

    6. Nope, not over the station. It's difficult to thread structure and elevators and such down through a rail terminal. The station, whatever it becomes, can stand on its own. What I am suggesting is that north of The Alameda, as the tracks swing through the area parking and warehouses around CEMOF, the tracks could be enclosed above grade and the surrounding development could climb up and over it. The site is huge, and if crossing above the tracks is planned, outdoor plazas could gradually elevate from street level in the east to up and over the tracks.

      The buildings along this park in Berlin are built over the subway line as it ascends to an elevated line. Before the building were built, you would see a two-track line ascending to an elevated line in the distance. The buildings were developed above the existing line.


  20. @Drunk: I see 4 blue (Caltrain) & 4 red (HSR) & 3 yellow (non-electrified) tracks.

    1. (Oops. I swapped red & blue; but the drawing shows 4 tracks for each.)

    2. I assume those 4 Caltrain blue lines are for platforms at Diridon. But the 4 HSR tracks continuing on north is what I can't quite understand. Perhaps they merge back into 2 tracks further beyond...?

  21. Finally some information on weekend service from the October LPMG agenda: " Weekend Service: Subject to additional financial analysis and budget confirmation, Caltrain expects to offer service every 30 minutes"

    1. Holy Crap! What is this? 1985? Fucking FUTURISTIC levels of service there, Caltrain.

      Now ... just remind us, yet again, just WHY THE FUCK THIS HASN'T BEEN THE ABSOLUTE MINIMUM LEVEL OF SERVICE AT ANY TIME OF THE DAY ON ANY DAY for every fucking day the last three fucking decades? Any reasons? Come on! There must be reasons, right? Right???? Right. Right.

      OH LOOK AT US AND OUR LPMG! WE'RE OVER-ACHIEVING! SEND MORE OPERATING SUBSIDIES! SEND MORE HALF BILLION DOLLARS FOR GRADE SEPARATING ONE SINGLE FUCKING ROAD CAPITAL FUNDS! (Also, we need to reduce service to once every two hours for each single road grade separation. Because a half billion dollars doesn't come cheap.)


  22. Metra wins $170m grant to “support” purchase of up to 16 BEMU trainsets

    “We think zero-emission trainsets could be an exciting and positive addition to Metra’s fleet for a variety of reasons,” said Metra CEO/Executive Director Jim Derwinski. “Beyond the environmental and noise reduction benefits, they also offer savings in energy consumption as well as better efficiency, flexibility, and reliability. Along with our other initiatives, they would continue to make Metra a responsible and effective leader in green energy technology.”

  23. Cost of SF Downtown Extension Swells to $8.2b

    The Portal is about to get much more expensive, according to an estimate presented at a Metropolitan Transportation Commission (MTC) meeting Wednesday.

    The proposed rail tunnel is intended to connect Caltrain — and future high-speed trains — at the Fourth and King streets station to the massive new Salesforce Transit Center. Project costs have ballooned from $6.5 billion to a staggering $8.25 billion since October 2022, according to MTC.

    The cost increase is attributed to a mix of factors, including inflation, expanding project timelines and a risk assessment from the Federal Transit Administration that determined that the prior cost estimate was unrealistic, according to John Goodwin, a spokesperson for the Metropolitan Transportation Commission.

    1. I guess it's time for higher bridge tolls and other revenues, though as usual you'll be competing with VTA and its BART to Silicon Valley extravaganza.

    2. Is this cost increase high enough to end the BART extension in DIRIDON, and cut Santa Clara station and yard from the Phase 1?

      BART could push for Santa Clara independently as Phase 2 with its own campaign for funding.

    3. The Portal project is in San Francisco, not San Jose. And the contractors for BART San Jose have shrewdly planned their project so that digging would start from the far end, in Santa Clara, precluding any reduction in scope. That said, SF and SJ are connected by a common aquifer of federal funding, and if one project pumps it too hard, the funding level will drop for the other. In my opinion, both projects deserve not to be built.

    4. Of course they want to start in Santa Clara. It's the part of the project everyone recognizes is the worst and needs to be cut first.

      At least Lendlease isn't joining the BART extension or high-speed rail contractors. (Yet? Maybe to help design and build stations?)

    5. InfrastructureWeak05 November, 2023 15:09

      That would be because Lendlease does building development and construction, not infrastructure construction.

    6. HINT: It's never too soon for a contractor to take public money for stations that may get built someday on a public project in addition to anything else along a right-of-way. Plenty of concept and other activity related to station design can be charged beforehand.

  24. Railway Gazette and ABB to present the following free webinar:

    Decarbonizing rail traction, a free webinar
    November 20, 2023 09:00 - 10:00 GMT

    Discover the innovative and cutting-edge technologies that are driving the decarbonisation efforts in rail, paving the way for a greener and more sustainable future. From electrification and renewable energy sources to state of the art rolling stock, this session uncovers the strategies and initiatives reshaping the traction landscape.


    Professor Stuart Hillmansen
    Professor in Railway Traction Systems
    University of Birmingham

    Max Linier
    Project Leader Diesel phase-out program rail
    Deutsche Bahn AG

    Edgar Keller
    ABB Traction Division

    Register free HERE

  25. VTA Orders Herrenknecht TBM for SJ BART tunnel

    VTA has ordered a state-of-the-art $76 million custom-built tunnel boring machine (TBM) to dig the largest single-bore transit tunnel in the world, nearly 54 feet in diameter. It's capable of digging 30 to 40 feet a day, according to VTA.

    "There's no turning back now," said San Jose Mayor and VTA Board Member Matt Mahan, in a statement released Monday. "This purchase cements our commitment to connecting the entire Bay via rail. When complete, BART to Silicon Valley will unlock tremendous transit capacity and economic growth for our region."

    The machine will be designed to operate in areas with high water tables, and the geology of sands, gravels, silts, and clays prevalent in the South Bay.

    VTA said there were extensive geotechnical investigations through comprehensive boring samples along the route that will connect the downtown San Jose Diridon Station and Santa Clara with the newly opened BART Berryessa station in North San Jose.

    "This is a momentous next step," said Santa Clara County Supervisor and VTA Board Vice Chair Cindy Chavez. "It represents more than just moving dirt — it represents being one step closer to moving people to jobs, to family, and to the entire region. Thousands of jobs will be created during construction, with thousands more fueling this transit-oriented development."

    The tunnel boring machine will be designed, manufactured, assembled, and tested by the manufacturer Herrenknecht at its German factory and then disassembled for shipping. The pieces will then be shipped to Santa Clara County and reassembled at the project's West Portal/Newhall Yard where the tunnel work will begin in 2025.

    The machine is expected to take approximately three to four years to complete the 5-mile-long tunnel. The boring machine will eventually be extracted on the east side of US 101 near Las Plumas Avenue in San Jose.

    The 6-mile BART extension will include 4 new stations in San Jose and Santa Clara. The cost has been estimated at more than $12 billion.

  26. At what point will the rollout of the new trains on this new schedule be interrupted by the months-long replacement of the truss bridge in Palo Alto?

    1. Over San Francisquito Creek? A plan is expected (hoped for?) by 2024-25.

      Meanwhile, repairs to the north bank are delayed, with an El Niño threatening more rain. This could get interesting, with or without Caltrain Super Sacks.

  27. Ben: bridge replacement is probably years off … and when it does happen, I think there’s a fair chance they’ll first drop in a pre-fab shoofly bridge on the upstream side to keep trains running across the creek.

  28. Wow! The UK’s “The Sunday Times” breaks huge HS2 cost fraud scandal:

    British high-speed rail project HS2 accused of huge fraud

    1. There has been bad news, unsurprising to some, with the state's high-speed rail project, but something really good remains to be seen. It's certainly possible, not just because it's another English country here, where trains are neglected the farther you travel and that helps certain "types" with public interest and scrutiny being largely elsewhere.

    2. Maybe this HSR guy has something to say, at the company giving us this work.

    3. Thank you for the link, I needed some cathartic screaming.

      I think I understand what went wrong. The problem was that they set out to "build a railway". They should have set out to "hold community meetings and, for a meaningful project legacy, build a park for wildlife to enjoy" instead. If they had set out to do that, then today they would have fifty aircraft factories.

  29. To help inform Palo Alto’s interminable grade separation design decision-making process, Caltrain presented its HSR blended corridor passing track alternatives and their pros, cons and underlying assumptions to the city council’s Rail Committee today. The bad news is that even after an assumed signal system overhaul allowing for 2-minute headways, Caltrain riders aboard trains being overtaken will need to suffer waiting at station platforms for at least 4 minutes.

    The full meeting video and accompanying slide deck are posted on the Rail Committee’s website: HERE

  30. Redwood City council and Caltrain last week jointly endorsed further study and development of an ambitious city-wide grade separation project that would eliminate all 6 remaining street crossings by elevating the Caltrain tracks and building a new, elevated 2-platform, 4-track midline multi-modal train and adjacent ground-level bus station that would allow express and future HSR trains to pass slower trains as per Caltrain’s board-approved long-term service vision:

    Grade separations at five (six!) crossings in Redwood City could cost up to $1 billion

    As shown in Figure 2 in the “Railroad Design Variances” subsection of the Project Summary Report on the city’s Grade Separation Feasibility Study website, 110 mph operation on the outside 2 of the 4 station tracks is said to require further unplanned for encroachment on some existing buildings.

    1. Can we just have Redwood City and Mountain View take over the grade separation process for the rest of the cities? They're both getting it as right as possible.

    2. Did you say four tracks? Don't forget San Mateo. See here and here.

      Oldies but goodies of note

    3. My pet peeve on the Redwood City grade sep is that project planners do not appear to be even remotely thinking about retaking Pennsylvania Avenue, which is a shitty back alley that fully encroaches on the Caltrain ROW. Car-brained planners are capitulating by not planning the critical southern RWC approach for four tracks, instead rolling out the excuse that there's no room for four tracks because of high-rise construction on the West side of the ROW. The fact that they fully own the east side of the ROW but don't plan on using it shows the depth of their windshield perspective.

      Pennsylvania Avenue should be erased entirely from Redwood City maps.

      Much more detail in my Redwood City coverage.

  31. BART trains became noticeably more crowded in September, when it started running shorter, 6-car trains on most lines.

    The strategy saves BART, which has recovered 45% of its 2019 ridership, roughly $12 million annually in power costs. But the change is aimed primarily at making riders feel safer on trains.

    The reasoning: People are less likely to commit crimes or antisocial behavior, such as using drugs, if train cars have more bystanders in them.

    BART says the shorter trains, paired with more patrolling police officers, has so far made an impact in reducing crime while improving reliability. The number of trains delayed by BART police responding to incidents dropped by 40% in October compared to May, when the agency operated 10-car trains.

    Unfortunately because Caltrain decided to order all of its 23 new EMU trains as practically inseparable 7-car sets, and as their current COO recently confirmed, they cannot enjoy similar power savings by running shorter trains during off-peak periods. Had it instead ordered the same number of cars as 40 4-car sets, it could much more economically run 4-car trains and 8-car trains (two sets coupled) only during peak demand periods. Using only one conductor per 4-car set would further reduce the cost of off-peak service, making it more economical to run more frequently and with later/longer service hours. Additionally, having 4-car sets would’ve nearly doubled the bathroom-to-riders ratio for added rider convenience on busier trains with higher bathroom demand.

    1. With 4-car sets they could’ve easily been able to couple/uncouple nose-to-nose in the field (as is not uncommon in Europe) from inside the cab with their automatic European scharfenberg couplers: youtu.be/qbzadTMp6vs

      So Caltrain squandered the whole value proposition offered by having EMU sets with scharfenberg couplers. 🤦🏻

    2. Caltrain is a Commuter Railroad. Baby Bullet Great Success!

      Commuter Railroads operate Big Trains at carefully calculated times (even if seemingly totally random if you don't get Commuter Railroad) and stopping at carefully calculated (ie random) collections of Commuter Railroad Depots in frantic Morning Rush and Evening Rush to The City.

      You need Big Trains for Rush, because Commuter Railroad, duh. Ask Caltrain's consultants: they are world renowned experts at Commuter Railroad. Best of the best. Caltrain wouldn't hire them if they weren't very very good at Commuter Railroad.

      Commuter Railroads store their Big Trains in Big Train Yards before Morning Rush hour, between Morning Rush hour and Evening Rush Hour, and after Evening Rush Hour. No need for little trains!

      Commuter Railroads store their Big Trains in Big Train Yards during the weekends. They're tired and need a big rest after five consecutive days of two Rush Hours a day.

      Big Trains also need to take long naps after each their busy Rush Hour trip to or from The City, so you need plenty of Extra Big Trains so they can cover for each other during well-deserved naps. Napping is very important at Commuter Railroads! Don't exhaust the Big trains.

      Good thing that we have experts at Big Earmarks and Big Cost Overrun and Big Project Delay so we can pay for all the Extra Big Trains we need for Commuter Railroad.

      I hope this explains why Commuter Railroad Caltrain needs lots of Big Trains.

    3. @anon you had me in stitches!

      @Reality Check, short trains were discussed here in the year 2017, then complained about in 2018 as Caltrain mulled an option order with a graphic of what should have been the optimal fleet. Sadly, Extra Big Trains for Commuter Railroad were procured at Big Cost.

  32. Endless Palo Alto grade sep drama enriched by 4-track Caltrain plans

    “Caltrain's new analysis of four-track segments adds a fresh complication to the city's convoluted and meandering quest to select preferred alternatives for grade separations (the redesign of the railroad corridor so that roads and tracks would no longer intersect). The City Council has spent the past decade evaluating alternatives for its rail crossings, gradually narrowing down its options from more than 35 to about half dozen.”

    1. There seems to be a lot of misunderstanding about what the 4-track Palo Alto segment is about and how it impacts their grade-separation planning.

      My understanding is that a short single-station overtake at either DTPA or CalAve or SanAntonio is needed for the "Moderate Growth" plan to provide a HSR overtake. With the "High Growth" plan a longer HSR overtake including all 3 stations is needed. Since the JPB guidance was to select the "Moderate Growth" plan and not take any action that precludes the "High Growth" plan, any grade separations implemented in Palo Alto must support future expansion to 4 tracks.

      I'm not sure how the selection of where the "Moderate Growth" overtake happens really affects the grade separation effort. They all need to support expansion to 4 tracks regardless.

      That said, the presentation to the Palo Alto Rail Committee was pretty interesting. Some points I noticed:
      - Considering only center platform FSSF for overtakes which is a change from the emphasis on SFFS outboard platforms in the past
      - Considering only type 20/24 turnouts with 50/60mph speed limits.
      - CalAve overtake impacts Peers Park. Apparently Palo Alto needs a "Vote of the People" for any impacts to parkland so this could be problematic.
      - Assumes any impacts to infrastructure e.g. San Antonio Overpass will trigger a complete rebuild.

      It seems to me like 3 tracks at Palo Alto and Cal Ave could be implemented pretty inexpensively,. so I wonder if it is possible to have the NB/SB overtakes in different places at least for the "Moderate Growth" plan.

    2. I’ve been told Palo Alto’s University Avenue station tracks are so widely spaced because there used to be a 3rd track between them.