31 August 2022

HSR Lays an Egg in Caltrain's Nest

California's high-speed rail project has finally reached a milestone 14 years after the passage of Proposition A in 2008 with the board's certification of the Final Environmental Impact Report for the San Francisco - San Jose project section.

This voluminous document has come a long way since the early days of a four-track 125 mph rail corridor initially envisioned for our region, having been whittled down to a two-track 110 mph "blended system" project that shares tracks with Caltrain without building any new overtaking tracks. Nevertheless, the cost estimates for the project have ballooned to $5.3 billion.

How do you add four high-speed trains per hour per direction, traveling at peak times between San Francisco and San Jose in just 48 minutes (by the way, 18 minutes slower than promised in the HSR bond measure) while building no new tracks and without disrupting Caltrain service? The answer is, you can't.

Instead, the high-speed rail project plans to take over the peninsula rail corridor and become a parasite to Caltrain. Here's how.

The FEIR's Volume 2 Appendix 2-C Operations and Service Summary contains what is described as an "Illustrative Timetable" of this blended service featuring 6 Caltrain + 4 HSR per hour per direction. The assumptions for this timetable include:

  • Caltrain is operated as a skip-stop service with 3 different stopping patterns.
  • Caltrain EMUs operate at up to 110 mph, their design top speed.
  • Caltrain station dwells are 30 seconds, consistent with system-wide level boarding.
  • Caltrain operates in salvos of three closely-spaced trains every half hour, leaving large gaps for high-speed trains to travel without being delayed by Caltrain. This bunching is flatly denied on page 3.2-91 of the FEIR, inconsistent with the Illustrative Timetable.
  • Two out of three Caltrains are held for about 5 minutes at either Bayshore or Lawrence station to allow high-speed trains to overtake them.

Additionally, we can make the reasonable assumption that peninsula commuters will be priced out of high-speed rail service, meaning that HSR will not provide many trips with origin and destination between San Francisco and San Jose. Every local peninsula trip taken on a high-speed train potentially displaces a longer statewide trip with much higher fare revenue for the operator, especially at peak times. In a yield-managed fare structure, local HSR trips will therefore be priced punitively. This isn't just speculation, it's well-established practice: Amtrak's premium Acela Express can technically be used to commute from Stamford, Connecticut into New York City, but it can easily cost you over $100 per trip, compared to the $15 peak fare on Metro North. In much the same way, every local SF-SJ passenger would potentially displace a more lucrative SF-LA passenger. Punitive local ticket prices will mean zero local ridership, so we shouldn't figure any HSR services into the scoring of service quality.

Let's put this in our handy taktulator and figure a score for this timetable. Remember, Caltrain's 2011 timetable is the baseline with a score of 100.

HSR FEIR Caltrain, no DTX: Score = 135 service points -- compare that to 147 service points with regularly-spaced 8 train per hour service, as planned in Caltrain's service vision and enabled by a new overtaking station in Redwood City.

HSR FEIR Caltrain, with DTX: Score = 235 service points -- compare to 250 service points with regularly-spaced 8 train per hour service as planned by Caltrain.

In both cases, the HSR timetable improves over the 2011 timetable (score = 100) by only 75 - 90% as much as the planned Caltrain timetable. Despite the higher train speeds, the bunched-up and irregular skip-stop service pattern with long overtaking dwells makes numerous Caltrain trips less convenient. On the plus side, this blended service can be operated with a fleet of 6 fewer trains than Caltrain's more frequent 8 tph service pattern. On the minus side, those trains will be more crowded. Note our taktulator tool measures service quality only from the standpoint of one typical user (weighted by origin and destination population and jobs density), without quantifying overall ridership demand or the resulting level of crowding.

Confronted by numerous stakeholders with the seeming contradiction between Caltrain's plans and its own blended service planning, the HSR authority offers Standard Response FJ-Response-GEN-4: Consideration of 2040 CaltrainService Vision and Caltrain Business Plan, with key points summarized below:

  • Caltrain's Service Vision is aspirational and isn't an approved or funded project.
  • The Service Vision is insufficiently defined to be analyzed in the EIR and is not "reasonably foreseeable" under CEQA.
  • The impact on Caltrain service wouldn't rise up to the level of a "significant impact" anyway.
  • The illustrative timetable is only used as a reasonable basis for analysis and there may exist better timetables.
  • It will be Caltrain's job to environmentally clear (and fund) future improvements associated with the Service Vision, such as additional passing tracks to support HSR service in the corridor.

In the HSR project's view, Caltrain will be fully on the hook for upgrading its own facilities to continue hosting HSR even as it becomes more difficult to do so. This sets up unhealthy incentives where capital projects that actually improve Caltrain service, and might create inconvenient "facts on the ground" for HSR, are quite likely to be delayed and de-funded to ensure the corridor stays clear for the future hatching of the giant egg that HSR just laid with this EIR.


  1. At least the EIR plan assumes level boarding for Caltrain. Maybe this will motivate them to finally start planning for it. The other nugget I got from the FEIR is that the 3 triack "long middle" overtake is by far the best option and was only omitted due to fear of PAMPA NIMBYs. Adding a 3rd track from Hayward Park through Mountain View should be the long term goal for Caltrain if they believe HSR will eventually come. Unlike 4-track segments, 3 tracks can in theory be done without needing any expensive new grade separations and fits easily in the existing ROW. 3 tracks with grade crossings is the plan for SJ-Gilroy so I don't see why it can't be for the peninsula too.

    1. @jpk122: SJ-Gilroy plans are only for 2 electrified tracks. The “3rd track” will be for UP and won’t be electrified. UP so far continues to refuse to run its freights under wires.

    2. @anon Whether the third track is electrified or not the safety concerns with 3 track at grade crossings are the same. If it is ok for SJ-Gilroy it should be good enough for SF-SJ.

    3. Hmm. Not sure how HSRA imagines being allowed to create 3-track crossings south of SJ without triggering CPUC and/or FRA into requiring grade separations. I know they pointedly don’t plan or propose funding any new grade seps between SF and Gilroy, so I’ll have to ask them about that if nobody here knows.

    4. @Reality Check: FRA doesn't seem to have any problems with Brightline creating new 3-or-more track grade crossings in future 110 mph territory (see video for an example). Any constraints here in California will likely come from CPUC, not FRA.

      @jpk122s: I shared your hope about level boarding for about a microsecond, and then I saw the joint letter (see page 35) recently sent to the PCJPB directors from Michelle Bouchard (Caltain executive director) and Brian Kelly (HSR CEO) talking up their collaboration on a list of corridor projects that includes DTX, SJ-Gilroy electrification, Diridon, grade separations, passing tracks, Millbrae, and the new Brisbane maintenance yard. Notably absent from their list is system-wide level boarding, which is necessary to reach the short Caltrain dwell times that enable HSR to operate on the corridor.

      My working theory is that the Caltrain ED and other decision-makers in the Caltrain organization (including their private-sector consultants) still think of level boarding as a nice-to-have ADA passenger amenity, and not the sine qua non requirement to increase train speed and reliability.

    5. Didn't HSR originally want 4 track, grade separation along the whole corridor. Caltrain should be saying that their plans are for a regular, frequent service and if HSR want to use the corridor then they should fit in. If they want faster speed then they should pay for the overtake tracks to achieve it.

    6. "Caltrain should be saying that their plans are for a regular, frequent service and if HSR want to use the corridor then they should fit in. If they want faster speed then they should pay for the overtake tracks to achieve it."

      "Caltrain" is simply a host body for feeding the consultant mafiosi chest-bursters.
      Nothing more, nothing less.

      If this were not that case, just once, just once, some time in the last twenty years, somebdy at Caltrain would have done something that improved service or cut costs. They haven't! This can't be explained purely by the (demonstrated and indubitable) rank unprofessionalism, stupidity, ignorance, incuriousity, and insularity of the people who "work" at Caltrain or those who "govern" it, because if that were the case then at least something not-bad would happen once in a while by pure accident.

      Batting 100% on choosing "maximal cost" and "minimum service" takes intent.

  2. There is no justification for running HSR on the Peninsula, because the only way to do it without permanently limiting both Caltrain and HSR is to institute through-running via a new Link21 transbay tube (as Clem convincingly argued last year in "Down the Tubes with DTX")--and once you accept the need to build that tube, the Peninsula becomes by far an inferior way to get to San Francisco, even without the numerous terrible decisions depicted in this EIR.

    The fastest, most direct route to San Francisco from the Central Valley is plainly Altamont-Dublin-Bayfair-Oakland-Transbay. This has the added benefit of serving Oakland along the way, which has much higher ridership potential than San Jose because it is much more dense and strongly centered (even the heavily sandbagged-for-SJ alternatives analysis which led to the Pacheco plan acknowledged this in absolute terms). The argument against getting to SF via Oakland was always the cost of building the tube, but if we have to build the tube anyway to prevent HSR from strangling Caltrain, that point is moot.

    This route would require far less eminent domain and new concrete than you might think, especially when accounting for other projects that are already necessary or in planning.

    A right-of-way from Lake Merritt Channel to I-238 already exists (Niles Subdivision til San Leandro BART, then Oakland Subdivision paralleling BART to 238). Capitol Corridor's vision plan floated the idea of acquiring land for passenger-only tracks through here, with little to no need for acquiring homes or businesses (I think there's one mini-storage facility that would need eminent domain). Oakland Sub is only lightly used by freight, and the northern part of Niles Sub is wide enough for four tracks if you take that mini-storage place.

    A fully-constructed two-track railway already exists along 238 and 580 from there to Dublin/Pleasanton (currently suboptimally occupied by BART). Re-gauging this line and putting up catenary wouldn't be nothing, but would be much cheaper than pouring all the concrete from scratch.

    Valley Link has already finished EIR for a partially-electrified (I know, I know) standard-gauge line from there to Tracy and Lathrop, including some improvements in the Altamont Pass. It's not funded yet, but eventually it will get built regardless of HSR.

    Even if we were only thinking about regional rail and not HSR, the whole Tri-Valley corridor ought to be all-mainline, without the forced transfer to a BART train that makes all local stops the whole way to Oakland and SF. Re-gauge the Dublin Canyon segment, delete the redundant and little-used West Dublin and Castro Valley stations, and run express with only the following stops from Altamont til SF: Livermore, Dublin/Pleasanton, Bay Fair (BART transfer), Oakland (BART transfer), Transbay (BART/Muni transfer). Because this corridor won't need as much regional rail as the Peninsula, blending with HSR will be much easier.

    The only totally new ROW needed would be through inner Oakland, perhaps closely paralleling 880 with a transfer station to BART at the foot of 980 (which should unequivocally be demolished), plus the connection to HSR in the Central Valley. Some kind of Oakland ROW and station will be necessary even with Clem's Peninsula+through-running plan. And of course building tracks from Madera to the mountains is necessary for HSR regardless of Pacheco vs. Altamont. Ideally Altamont Pass and Dublin Canyon would get new tunnels to allow higher speeds, but I think together they would be less tunnel mileage than Pacheco (also maybe wouldn't have to be finished to open the line on day 1). So none of these things are really added costs above and beyond what would be necessary to make the more circuitous Peninsula route work.

    1. The only truly new costs of my plan, above and beyond the status quo of just building Peninsula HSR, Link21, and Valley Link, are:
      * Acquiring freight ROW from Oakland to Bay Fair, which is either little-used or has room for four tracks
      * Building the tracks on the existing ROW from Oakland to Bay Fair (roughly 10 miles), which would require some grade separations
      * Re-gauging and electrifying the BART Tri-Valley line (roughly 13 miles, or substantially less if a Dublin Canyon tunnel is built)
      * Electrifying the non-electrified parts of Valley Link (lol)
      * Acquiring a mini-storage facility on Oakland

      A couple weaknesses of this plan, off the top of my head
      * HSR trains from Transbay would have to through-run down the two-track San Francisco segment of Caltrain to get out, maybe to an HSR yard around Bayshore Caltrain. If HSR trains go out of service at Transbay, as I think they should, this should be fine--even if it's running right behind Caltrain, the out-of-service HSR can just go slowly and wait behind Caltrain at stations, so there's no need to operate Caltrain with a weird bunched service pattern to make up for the lack of passing tracks in this segment. But it might be hard to continue revenue service for some HSR trains through to, say, Millbrae/SFO without constraining Caltrain. Maybe it could be done off-peak.
      * Does not serve SJ in the initial system. This is far less important than serving SF and Oakland (and doing so as well as possible), because those cities have much higher ridership potential. And the current plan's schedule (among other things over the years) makes plain that strict adherence to the service pattern laid out in the original bond measure is not required. Also, since CHSRA somehow managed to blow the entire budget in the Central Valley, a new bond measure might end up being needed, which could override the old one. Still, it is a missing feature compared to the status quo plan, even if it's a worthy tradeoff. SJ service could and would be added later: as Alon Levy has noted about the Northeast Corridor, once there's compelling service between the most important destinations, it becomes increasingly attractive to add extensions to secondary destinations. Maybe a line down 680 from Dublin/Pleasanton to roughly the location of the last SETEC tunnel could send trains onto the Warm Springs Sub, cross over to 880, end up at Diridon?

    2. This is just bad fanfic. Dublin/pleasanton is never going to be regauged and a freeway median is a poor place to run a HSR line anyway. It needs to get up to line speed as soon as possible when departing the bay to achieve reasonable times to LA, crawling along a freeway median is a nonstarter. Your proposal requires a new transbay tube and ~50km of new right of way, much of it tunneled. That's going to be 10s of billions (charitably) just to get service of the ground. Not going to happen.

      Altamont remains the only reasonable way to get HSR into downtown SF, at achievable costs and with reasonable speeds and operational constraints within this century. Gets HSR up to line speed shortly after departing redwood city, serves oakland and SJ with a BART transfer at a new station near lake elizabeth. It's the only real option.

    3. @Mobert Roses while what you say about Altamont is mostly true, at this point we have an approved EIR for the Pacheco route. It may or may not ever actually happen, but it is now far more likely than any alternative. All it takes is a bit of political will in the Governor's office to find the $$ to get it done. Whether you approve of the chosen route or not, I think it's important to acknowledge the huge effort required by CAHSR to get it to this point.

    4. Crawling? The freeway through the Tri-Valley is arrow-straight and already impacted noise-wise, a.k.a. the perfect place to run HSR through a populated area at speed. Like I said, the curvy Dublin Canyon part ought to get a new tunnel. Of course the east bay part wouldn’t be at line speed, just like the peninsula route, but my east bay route is both shorter and straighter, so actually my plan would get up to line speed sooner, and would slow down less while in the populated corridor, in addition to being shorter overall.

      Again, through-running via the new tube is a necessity anyway, to avoid permanently constraining both HSR and Caltrain. Clem made this point really comprehensively in “Down the Tubes with DTX” last year. So it’s not an added cost beyond what the current course of action will ultimately require. If you don’t buy the need to build a new tube, then I agree that my idea is a lot less compelling, but I just don’t see how having Transbay as a terminal station won’t constrain the hell out of both systems even if improvements are made on the rest of the corridor.

      My plan also doesn’t require more tunneling than other options. The standard Altamont route needs a tunnel through Altamont Pass and then a second one through Sunol Canyon-ish to reach Fremont. My route basically just replaces the Sunol Canyon tunnel with a Dublin Canyon one. Either the Sunol or Dublin option probably requires less tunneling than the Pacheco status quo.

    5. @jpk122s I’d be more inclined to agree if shovels were about to go into the ground and it was a choice between building Pacheco now or delaying for years and maybe getting nothing at all. But the truth is there’s no funding on the horizon for any of this, and no sign of that political will. That’s depressing, but it also means there’s time to get it right.

    6. @Bryan Anderson: unless something or someone manages to change their minds, Valley Link staff and board still remain hellbent on running Stadler FLIRT (or similar) HEMUs fueled by a green hydrogen plant they want to build on an available lot in the western outskirts of Tracy. They imagine their H2 plant could also fuel LAVTA and other agencies’ future H2 bus fleets.

    7. @Anonymous Oof that’s unfortunate. I also looked back at their DEIR and it looks like their BEMU with in-motion charging option only proposes catenary in the Altamont segment (which HSR would bypass with a much straighter tunnel anyway). I get that they feel they have to avoid it in the shared Tracy-Lathrop segment, but not putting it up in the Tri-Valley segment is unfortunate. Also their Livermore station placements are terrible, and their station designs are generally surface-parking hell. And don’t get me started on their Dublin/Pleasanton design which requires you to go downstairs and EXIT THE STATION just to enter the BART station and go back upstairs. Sigh.

      I wonder how much of this is fixable without totally restarting CEQA. Maybe a sane BART transfer station design could be argued to not have substantially different impact from the stupid one (except higher ridership, lmao). Obviously I think there shouldn’t be a transfer at all because I want the BART line for mainline rail, but if there is, it should be an eBART-style cross-platform transfer. Also they did analyze a full OCS alternative “in less detail”, so maybe they’d just have to flesh that part out in more detail if they could be convinced it’s the right way to go?

      Oh also, ACE is independently proposing an Altamont tunnel which would be used by both ACE and Valley Link and would allow speeds up to 125mph. Higher than that would be better obviously. The design doesn’t seem to be far enough along to say whether 125mph will be a hard limit imposed by track geometry or simply the top operational speed of ACE. Does anybody know what the proposed top speed is for the Pacheco tunnel?

      Anyway—bad decisions aside, there’s about to be a double-track freight-free rail line down the arrow-straight 580 corridor to the Altamont Pass, possibly followed up with an Altamont tunnel. Somebody ought to do something non-stupid with it.

    8. Ugh I do have to take one part of that back—valley link will be mostly single-track through the tri-valley, so it would need to be double-tracked in my plan. Some curve straightening would also be desirable east of Isabel, but isn’t being done as part of valley link. We really can’t have nice things.

    9. @Bryan Anderson my understanding is that there is insufficient vertical clearance for OCS in this segment due to freeway overpasses. I think there are more double tracked segments now than originally proposed as BART insisted on 15min frequency so as not to overload their trains with transfers but there are still single track segments,

    10. I see in the DEIR that they claim regrading would be needed to support OCS, but they didn't quantify how much. Interstate bridge clearance must be 16 feet. A British (Network Rail) guide[1] states that 4780 mm (15' 8") clearance is needed from top of rail to bottom of bridge, provided that the OCS is fitted to the bottom of the bridge, with a bit less allowed in special circumstances, and 5100 mm (16' 8") for free-running OCS.

      CAHSR's OCS requirements document[2] states that free-running, which presumably enables higher speeds (although I didn't find concrete numbers), is to be used unless this is not possible in a shared corridor. The document states that 27' clearance from top of rail to bottom of structure is required for free-running and 24' for fitted. It also states that the contact wire will be placed at 17' 5", whereas the British guide puts the minimum contact wire height at 4.165 m (13' 8") and the default at 4.7 m (15' 5").

      It's not clear to me whether these discrepancies are due to speed (the British guide is not HSR-specific), loading gauge (smaller in Britain), or reflexive conservatism on CAHSR's part. So it's hard to say how much lowering would be needed: possibly very little if nothing faster than Valley Link ever runs on it, or maybe quite a bit more to support HSR.

      And good, the more double-track the better (except at the transfer station where there ought to be one terminal track on each system at the same platform, like eBART...)

      [1] https://www.bathnes.gov.uk/sites/default/files/sitedocuments/Planning-and-Building-Control/Planning/nr_a_guide_to_overhead_electrification.pdf

      [2] https://hsr.ca.gov/wp-content/uploads/docs/programs/eir_memos/Proj_Guidelines_TM3_2_1R01.pdf

    11. If you're regrading to add rail in the median of the, you'll be moving all sorts of drains and pipes, so while you're at it, JUST GRADE THE RAILBED LOW ENOUGH to provide standard clearance for the OCS at over crossings.

    12. Re-grading (lowering) the I-580 median for Valley Link track(s) for OCS clearance under existing overpasses will likely also require failsafe pumps (with backup generators) to prevent them from flooding. Yay!

    13. 1. If you're quoting any standard from the English-speaking world, it is either completely irrelevant (UK, Australasia) or BAT SHIT INSANE NEANDERTHAL NONESENSE (USA, Canada) and likely just pulled out of somebody's ass, at huge extra cost (anything made-up by the rent-seeking know-nothing criminals at WSP doing business as the fictional "public" shell corportation CHSRA.) Maybe there's something to learn from India, maybe.

      Anyway. Stop it. Just because it's google-able doesn't mean it means anything.

      2. If you're talking about doing anything in freeway medians you are BAT SHIT INSANE already, and carrying on about freeway overpass clearances or whatever is meaningless. Just stop it.

      Anybody at any "public" agency talking about transit in freeway medians needs to be put out of our misery, immediately.

      Just stop it.

      3. "will likely also require failsafe pumps (with backup generators) to prevent them from flooding"

      Oh, we're talking about Caltrain's BAY SHIT INSANE "split" grade crossings and hideous proposed excvation of freeway-wanna-be urban "expressways" underneath at-grade Caltrain tracks now, are we? And how about those fatalities in in Millbrae earlier this year? Death is too kind a fate for those involved.

      Again, these people need to be put out of our misery. All of them.

      Just stop it!

    14. I'm not at all fixated on the median--it's worth getting creative and thinking about where it makes sense to parallel the freeway (vast majority doesn't have adjacent residential), use elevated structures, or use the median. Following freeway ROWs is not unheard of in continental Europe (I'm thinking of the LGV Nord-Europe but I think there are others) precisely because they are already-impacted corridors that usually go more or less directly between major nodes (in this case Oakland and the Altamont Pass, the only sane crossing of the mountains). Separately, SETEC thought the way to do the Peninsula was to use the US-101 ROW. So maybe following a freeway is not always a "BAT SHIT INSANE NEANDERTHAL" English-speaking-world idea.

      Fun fact: terminating at Transbay is so bad that it's still a hard limit of 12 trains per hour even if you use SETEC's 101 route to avoid sharing tracks on the Peninsula. Oops!

      580 is a mostly straight corridor that makes a beeline for the most direct route to San Francisco, which seems silly not to consider *if* a transbay tube has to be built anyway. I am aware you think it does not. And if it does not, then forget it, this route isn't worth it. The expense of the tube was the (valid) reason that this route wasn't seriously considered the first time by Altamont advocates. But here we are stuck with an inadequate terminal station that will permanently constrain both Caltrain and HSR. I'd love to hear other options for dealing with that that don't involve either sticking Caltrain with terrible, inadequate service patterns or leaving half of the HSRs in the boonies of San Jose (lol, no one will ever ride the 4 trains that terminate at Diridon, they'll be canceled asap).

      If you have to build the tube anyway, why take the long way?

    15. Following freeway ROWs is a relatively common practice and not bat shit insane, running in the median is.

      Richard has made compelling arguments in the past about the lack of suitable ROWs in the east bay. The penninsula ROW already exists and CAHSR needs to serve SFO. Through running transbay at this point is a tall order

    16. Fine, forget the median and run alongside. My point isn’t about the median, it’s that suitable ROWs in the East Bay do in fact exist. The almost entirely straight freeway corridor through the Tri-Valley, combined with portions of the Oakland and Niles subdivisions, provides a quite direct path from the Central Valley to Oakland that doesn’t involve crayoning through a bunch of existing development or boring 30-mile tunnels.

      It’s also a corridor where we know (from all the freeway use) that there’s a lot of regional travel demand, but where that demand is currently served very inadequately by ACE and BART. A regional express service with 200 km/h top speed (running Transbay-Oakland-Bayfair-Pleasanton-Livermore-Tracy then branching to Lathrop-Stockton and Manteca-Modesto) would take a ton of cars off the road, in addition to being the fastest way to get HSR to the two city centers where people are actually trying to go (that is, not San Jose).

      Does HSR really need to serve SFO directly? HSR stations at major airports do exist internationally, but aren’t the norm even in countries with good HSR. I am skeptical that demand for HSR to SFO (which must necessarily come from smaller cities in the Central Valley, since nobody will take HSR for short distances and nobody from LA is going to take a train for multiple hours just to fly out of SFO) will hold a candle to demand in city centers. At any rate, it might be possible to through-run a couple HSR services per hour to SFO, and if not, a fast same-platform transfer to a Caltrain express is not that bad.

      Clem laid out exactly what it would take to through-run Transbay last year in this post and its comments, and while it wouldn’t be cheap, it could be done, and the supposedly show-stopping eminent domain would not be the end of the world: https://caltrain-hsr.blogspot.com/2021/09/down-tubes-with-dtx.html

    17. The clock is ticking. I see the Merced and Bakersville endpoints being under contract by this time next year which would imply a big push for the next focus by the end of next year, my guess electrification of Gilroy to SJ.

    18. @Bryan: you questioned whether HSR really needs to directly serve SFO? As evidenced by the need for a transfer, Millbrae ≠ SFO.

    19. @les: I don't believe any funding for SJ-Gilroy has been identified, has it? Finding it is going to take a political push led by Newsom, who has shown a preference for completing the Central Valley segment and deferring the Bay and LA segments. I have serious qualms about this strategy (opening a near-useless HSR system with no extensions under construction will probably make voters dismiss it as a boondoggle, rather than get excited to fund more of it), but a silver lining could be more time to get it right, if someone with political influence wants to push to get it right (I do not have a ton of hope).

      @Reality Check: Correct, of course. The long shadow of the botched BART-to-SFO. Nobody wants to use the Millbrae transfer to BART-to-SFO, which is all the more reason why HSR serving Millbrae is not the most important thing.

    20. Funding for finishing the 171 CV segment has been secured (4.2bil prop funds and C&T 1-2bil; still waiting on word for the 1+ billion from Fed Infrastructure grants to electrification and buy trainsets. Will know in a few months). Contracts for engineering have been out for some time now.

      MTC has been floating 7+ billion for SJ and rail south for some time now. I can't see any other segment gathering the political attention SJ - Gilroy is. Not my first choice but the choice with most interest.
      "Boris Lipkin, high-speed rail’s San Jose-based Northern California director, said the MTC allocation would be applied to electrify the line south from Tamien to Gilroy"

    21. There is (and always has been) a sensible solution for the BART/Caltrain/HSR/SamTrans connection to SFO. Abandon the BART connection and extend AirTrain all of 1.5 miles over 101 to Millbrae Station. Trip time from the last airport terminal to Millbrae about 3 minutes. The southern portion of the BART wye could be repurposed for the overcrossing. It will never happen, because politics.

    22. surprisingly feasible too, if the airtrain rolling stock can run in the tunnel. airtrain can take over the unused eastern platform. you'd have to single track through the tunnel but for a ~1km single tracking section at the headways BART is running to milbrae that shouldn't be an issue.

      Quentin Kopp should be held liable for the capital cost.

    23. @Marc: yes, many of us transit advocate/wonks, including the Dennis Sammut (San Bruno’s Artichoke Joe’s owner) funded “Coalition for a One-Stop Terminal” (COST) group, fought the Kopp-led BART/Samtrans cabal over BART into SFO vs. a single, consolidated BART/Caltrain/HSR/Samtrans transfer terminal on the existing Caltrain line conveniently served by the free-to-ride SFO AirTrain providing all transit users with a single free transfer to/from all SFO terminals, rental car center, etc.

    24. If they felt the need (of course, they would, just like the OAK connector), they could charge $5 to enter/exit AirTrain platforms at Millbrae, it would still be better than the existing options. It would also reduce BART headways at Millbrae as the former SFO trains would terminate there. Perhaps two lines, clockwise and counter-clockwise around the terminals, and two AirTrain platforms at Millbrae, one train unloads/loads while another traverses the connector. Or splurge, and build out the connector with dual trackways for reliability. I can dream...

    25. Bryan Anderson,

      No, no, another transbay tube makes far less sense than rebuilding the dumbarton rail bridge. Less expensive option first unless something makes it implausible.

      Which is another reason I proudly voted no on HSR in 2008 as a centrist Democrat....everything I thought would go wrong with it has gone wrong.

      btw 10 trains per hour per direction as Clem describes above says there would a train every six minutes. Does anyone here really think that is practical?

      and DTX....lmao never going to happen. Surprised it's not cancelled already. There is already a light rail right across the street that does pretty much exactly the same place and would take the same amount of time lmao. Severe waste of money. Just hang a wire and ferry people across on a gondola, seriously.

  3. Seems like sending a third or half the HSR trains from San Jose up to Oakland would be worth considering. Never shoulda stuck the sole northern HSR terminus on a peninsula.

    1. San Francisco is THE terminus, but most if not all traffic should be routed through what has always been the real transportation hub of the Bay Area, Oakland. Despite what some say, it's not San Jose.

      Re-routing the high-speed trains to San Jose -- yes, they should go through Oakland -- is just to appeal to some little South Bay egos with a complex about their main city's size and importance. That also was behind the lunacy of choosing Pacheco over Altamont, to ensure trains went through San Jose. Never mind that the trains now go away from where most of the people are, including nearly all now and in the future, too, that would be those commuters the project is now invoking to defend the project (and promote it very weakly).

      And if that's not enough, now the project isn't even trying to minimize tunneling as professed earlier, but seeking instead a 13.5 mile tunnel under Pacheco Pass. (No doubt some fools will boast of that or claim bragging rights for the Longest Tunnel in North America the shiny trains use just outside of Silicon Valley.)

      [sigh] Contractor bloat and worse

    2. San Jose is the biggest city in the Bay Area so of course any trains should stop in downtown San Jose. I mean really the CORRECT way to do HSR would be to start it in Tracy and run it down the median of I-5 down to Los Angeles. Then people would get to it via Capitol Corridor, BART, etc. In that case there would be trains from SF that diverted across a rebuilt Dumbarton Rail Bridge to connect to Capitol Corridor tracks.

      I know, I know, I'm making too much sense.... the route was political motivated rather than being driven by common sense and the market, so of course it goes down 99 and not I-5.

      btw the CAHSR business plan says nothing about any kind of stop in Oakland. You guys should have voted no.

      I'm very proud I voted no in 2008 as a centrist Democrat.

  4. @Bryan Anderson: Yes, in addition to Altamont Pass being the better route for regional and longer-distance (inter-city) trains, Dublin Canyon is the better route to use, along with the I-580 (old U.S. 50) route directly, repeat, directly across the valley, reducing the distance across the Diablo Range and minimizing the part that is rugged and challenging terrain. It's better than what would be used, something like the SETEC route that would also incorporate a revived Dumbarton bridge crossing. Ignore those who gripe about a freeway alignment or "worse," a freeway median placement; that's just activism substituting for sound direction. Nobody normal wants to detour through downtown Pleasanton and Livermore between the East Bay flats and the Central Valley and beyond. Sadly, the route is taken already (though BART has still failed to reach Livermore, near the edge of the Bay Area there, in its campaign with VTA now to compensate South Bay people with their little city complex) and the overpasses may be trouble. It's a shame, including loss of the obvious station site at Dublin, crossroads since the terrain was formed and especially since that part of the Bay Area grew after World War II. Most consider something like the SETEC route instead now, the southeastern tangential route Peninsula to Manteca wye, with a revived Dumbarton bridge crossing of the Bay to the Peninsula. A new Altamont route should be support regional service, the entire heart of Nor-Cal through Sacramento, not just high-speed inter-city service. (The same is true for Burbank-Palmdale near L.A..)

    A future Bay crossing to Oakland and having high-speed trains use that route and the old WP Subdivision (BART East Bay viaduct) route (the crossing would serve Sacramento service as well) is obvious, but out of reach currently. (Neither MTC nor local politicians can be trusted.) Note the Bay crossing doesn't have to be a tunnel or tube and this is foolish to assume as so many do; a bridge would be better and permit transit roadways to be added, plus rec trail use for added fun. Oh, well.

    As with the high-speed rail project, so with local Bay Area transportation (where is the 238-380 Bay crossing, and maybe another near Alameda, both sides for rail and transit as well?) and demented activism with multiple subjects instead of rational, sound decisions. It's another loss.

    1. Agreed on most points, but I’d point out that agency turf is not immutable. A creative solution could involve putting BART as the agency in charge of mainline regional rail in Northern California. They’re already the “managing agency” for Capitol Corridor, and Clem’s proposal for merging Caltrain with BART is sound. If the resulting regional service in the 580 corridor would “still be BART”, maybe the agency would jump at the opportunity to replace the L line (L for Livermore lololol) with the kind of mainline that we’re talking about, as it would extend the agency’s empire rather than reducing it.

    2. Well, there would be many who would want some other agency "swallowing" (subsuming) BART, with a standard-gauge-oriented main world view, typically with overhead electric power, called something like, yep, Bay Area Rail Transport (a new "BART").

      The Dublin Canyon route is superior to the Mission Pass route even with the latter combined with a new Dumbarton bridge. (There is no reason for a tunnel there, even less than there is between west Oakland or Alameda, and San Francisco.) The problem isn't so much the broad gauge and removal of third rail but overhead and other clearances.

      The Tri-Valley constitutes much of the Diablo Range crossing on that route, meaning open, flat land, plus a direct route toward Altamont Pass using the I-580 alignment.* (Contemporary warped activism, which would want trains through downtown Pleasanton and Livermore, is always distinct from and should be excluded from transportation and transportation planning.)

      The Bay Area there reaches, as always, to the mountain crest and watershed boundary, for this purpose at Altamont Pass. (There are other passes in the vicinity, too.) It's freakish as well as exemplifying the state's decline when people with an outsider mentality speak of only the Bay flats and the City as the Bay Area, actually say the interior portion isn't, or that people come to the Tri-Valley "from Fremont and the Bay Area," which I have read before. They don't know the area they're talking or writing about. I fear the same gross incompetence lies in much gov transportation and other decision-making all too often and is getting worse, if anything.
      * The open, flat portion the Tri-Valley provides as part of crossing the Diablo Range, aided by Dublin Canyon or known ways to the south to reach it, too, is another reason, in addition to going where the people are and will be in the Central Valley for regional as well as inter-city travel by all kinds of new trains that could use the route, why Altamont Pass is so much better than Pacheco Pass.

    3. The thing is, I’m not sure which agency in the region with a mainline-oriented worldview would be better. Caltrain has shown that it still can’t conceive of modern mainline rail in any useful way (botched electrification, signaling, and rolling-stock acquisition, still no plans for level boarding, general incompetence much higher than BART’s in both operational and capital areas). BART already runs Capitol Corridor, and seems to do at least as good a job as anyone around here at running mainline rail (quite the low bar). Capitol Corridor’s vision plan shows that they mostly understand the basic elements of modern regional rail—they just don’t have any funding to implement any of it. So I think BART would do fine running regional rail, or at least no worse than anyone else around here. Of course the real solution is to hire capable people from abroad to run every aspect of Northern California transit.

      Fully agreed on the superiority of 580-Dublin over SETEC-Dumbarton, of course (and that Dumbarton should not be a tunnel under any circumstances). Clearances are annoying but we are talking about a few feet of regrading here, not blasting through a mountain range or even digging a full open trench. And SETEC would require a lengthy cut-and-cover through Fremont! It really is not close.

    4. I'm not the harsh critic of Caltrain that others are; I see it as anything on the Peninsula putting people somewhere within at least a long walk of the downtown area being better than nothing.

      Dumbarton revival (a bridge, why on earth a tunnel without any need or reason for it?) is separate from modern Altamont Pass rail crossing use. The interesting thing about SETEC and Dumbarton is that the route in question between Niles Canyon and Mission Pass, to and through Altamont Pass, to the Manteca natural wye in the Central Valley, with or without the Dumbarton extension to the west, is the farthest south any high-speed rail route should g, and coincidentally it is "sideways" or roughly perpendicular to the main direction of the oblique (not north-south) main axis of the state and travel along its main dimension. (Rotate a map to see. You'll also find yourself being kinder on Bakersfield-Palmdale, though not on Palmdale-Burbank, and Grapevine remains better. The Highway 99 route is also inferior for speed to the I-5 route for high-speed rail, but of the three major route decisions and blunders, this is the easiest understood and accepted. Understood but never accepted is the Pacheco Pass selection.)

      SF-Oakland, then south to the Altamont-Dublin Canyon route, is the better way. Note that this east-west route's extension is the highway connection, I-238 to I-380, the most important missing rung in the ladder of Bay crossings to build, could include rail or BART with it (and more). Another bridge north of that is needed, too, separately from a new rail-BART or transit crossing between Oakland or Alameda and downtown S.F. or close to it.

  5. Clem, what are you talking about with 115 mph.

    Currently the track speed is 79 MPH and CalMod does not change this.

    Theoretically, if we were able to eliminate all grade crossings on the route, trains would still not average 115 mph. Currently they allowed to go 79 but average 55+ at best (hence why baby bullets take about an hour).

    If the max speed were raised to 115 I would guess the average would be like 80 which is an improvement but not 115 average.

    Long before then, Joby and others will be running eVTOL aircraft from SJC to downtown SF and/or SFO and getting there in 5-10 minutes, on electric power (batteries/H2).

    Anyway let's deal in reality not what the track speed is or will be.

    1. Grade Xings aren’t slowing trains and won’t prevent HSRA (and maybe someday Caltrain) from running 110 mph through quad gate protected Xings.

      And Joby eVTOL flying cars? Don’t be silly!

    2. I refuse to believe that you actually just said "let's deal in reality" with a straight face in the very next sentence after talking about electric flying cars.

    3. Don't forget that those are supposed to be fully automated magical flying cars, like those fully automated magical EVs we're already supposed all to be using.

  6. Is it possible to have HSR pass Caltrain on the other pair of tracks? NB HSR could pass a NB Caltrain on the SB tracks while Caltrain loads/unloads passengers at a station. I’m not really sure where the upper bound on capacity lies for the 2-track corridor, and how different it might be with an ideal PTC system. Would be fun to play with this on the taktulator.

    1. Yes, but...

      You would need high speed crossovers available so the train would not need to reduce speed to get to the other track. You would need no opposing traffic for the time it takes the HSR to gain on, overtake, and pass clear of the stopped Caltrain, which is probably a long enough interval to restrict headways of opposing trains. AND MOST IMPORTANTLY you would need to adhere to a very tight schedule or the smooth overtake would fail, resulting in a cascade of delays. So realistically, on the Caltrain corridor, it wouldn't work.