28 January 2009

Judging a Train By Its Nose

One of the recurring themes emerging from public comments about HSR on the peninsula, at scoping meetings or on community bulletin boards, is the notion that HSR on the peninsula merely duplicates existing Caltrain service at astronomical expense and disruption to communities. The reasoning goes like this: why bother with HSR running at "half speed" on the peninsula if passengers can just catch a Baby Bullet in San Jose that whisks them into San Francisco? We've already got a bullet, so why do we need another one? If HSR is only running at "half speed" why not use the existing tracks? Those are good questions.

One thing is certain, Caltrain has scored a marketing home run with their Baby Bullet brand, borrowing the aura of high speed rail and pointy-nosed aesthetics to convey an image of speed and efficiency. Unfortunately, there are a couple of problems with confusing the Baby and its namesake.

Myth #1: Speed

"We already have a bullet. We don't need another one!"

Baby Bullets look sleek, and they impress when they blast by, horn blaring. However, they are nowhere near as fast on average as the proposed high speed trains. Timings from San Jose to San Francisco:
  • Caltrain local: 96 minutes - max speed 79 mph - average speed 29 mph
  • Caltrain Baby Bullet: 57 minutes - max speed 79 mph - average speed 49 mph
  • HSR: 30 minutes - max speed 125 mph (a.k.a "half speed") - average speed 94 mph
HSR at "half speed" is still nearly twice as fast as the Baby Bullet. Suppose HSR were terminated in San Jose, and that passengers transferred to Caltrain. Accounting for transfer time in San Jose, the effective speed of the Baby Bullet ride would approach three times slower than a single-seat HSR peninsula ride. The overall trip from LA to SF would increase from 2:38 to 3:15, an increase of about one quarter. If we compromise that much on the peninsula, why even bother with 220 mph in the central valley? Indeed why bother with high speed rail at all?

Ridership is very sensitive to total trip times and changing trains. Every second counts. That is the reason for terminating HSR in San Francisco, which has a large catchment area of potential riders.

Myth #2: Track capacity

"If HSR is running at half speed, let them just use the existing two tracks!"

Mixing trains with different average speeds on the same track reduces track capacity, measured in units of trains per hour (tph). A slow train must be given plenty of time to clear the tracks ahead of a fast train that follows; otherwise, the fast train will catch up to the slow train and get stuck behind it. While modern signaling systems allow 15 tph when speeds are homogeneous, today Caltrain can only manage 5 tph at rush hour because Baby Bullets and locals operate at very different speeds. Add HSR to the mix, and what little track capacity we have would collapse.

More tracks are needed along much of the peninsula to allow trains with vastly different speeds to overtake each other, thus freeing track capacity. This ultimately has benefits for local Caltrain service as well: with Baby Bullets allowed to operate on the new HSR tracks, local track capacity will increase, enabling more rush hour service to under-served stops like California Ave or Belmont or South San Francisco.

Never judge a book by its cover. Likewise, never judge a train by its pointy nose.


  1. 1. 30 minutes from San Francisco "minimum speed, minimum capacity, finest US rail professional engineering" Transbay to San Jose Rod "Not Yet Dead" Diridon Memorial Four Level Pangalactic Intermodal Rocketport is and always was a fantasy and and engineering fraud given the known characteristics (demographic and physical and self-inflected injuries) of the corridor.

    The average speed of HSR trains on the Caltrain line will and in fact must be almost exactly the same as the average speed of Caltrain expresses, and Caltrain expresses aren't going to be averaging over 100mph.

    That's how it works everywhere else in the world, and that (minimizing the number of different traffic classes, maximizing intensity of traffic per track, minimizing capital and maintenance expense) is the only way it can work on the Caltrain corridor. (And it is and will remain the Caltrain (= S-Bahn/RER/Overground) corridor: 80% Caltrain, 18% HSR, <2% freight/dinosaurs. Tails don't wag dogs.)

    Come on, Mr Caltrain-HSR, you already know this!

    Caltrain may speed up in the unlikely event that our friends at PBQD-VTA-BART-MTC don't completely defund all improvements in the corridor (you will be able to change to BART at SJ Rod "Still Breathing" Diridon memorial station for a trip via Warm Springs, Fremont, West Oakland, The Inconsequential City Whose Name Must Not Be Mentioned, Daly City and Colma to Millbrae Quentin "Unindicted" Kopp Memorial $1.8 Billion Intermodal Station), but HSR trains on the same tracks in the same corridor won't and can't and shouldn't operate at significantly different speed from that at which the fastest Caltrain services do.

    In fact, in an advanced (ie Translink-free) industrialized first world democracy you'd quite likely see HSR trains run in place of some regional services, accepting local fares for local trips in order to minimize the number of empty seats ferried between SJ and SF.

    Just for example: 06:02 (ICE); 06:34; 07:02 (TGV); 07:34; 08:02 (ICE); 08:34; 09:02; 09:34; 10:02 (ICE); 10:34; 11:02 (EC); 11:34; 12:02 (ICE); 12:34; 13:02 (EC); 13:34; 14:02 (TGV); 14:34; 15:02; 15:34; 16:02; 16:34; 17:02 (ICE); 17:34; 18:02 (TGV); ...

    And in our particular case, after all, most everybody in the state will be heading to or from San Jose, so it makes sense to try to fill the otherwise largely-empty HSRs from peripheral locations north of Santa Clara with workers and shoppers for SJ's throbbing CBD before all train seats are sold out to Silicon Valley titans heading off to do deals in Los Angeles or Los Banos.

    2. If is most definitely not the case that "More tracks are needed along much of the peninsula to allow trains with vastly different speeds to overtake each other."

    In reality the only place that Caltrain needs such tracks are between Redwood City and San Mateo. Much anything beyond that is either (a) contractor profit maximizing gold plated boondoggling (say it isn't so!); (b) dismally incompetent engineering (ditto!); or (c) a scare tactic used to inflate costs to such an extent that the project is infeasible and is cancelled (with the diverted funds reprogrammed to backfill unexpected cost overruns in the BART SJ extension, itself defined as a Vital and Essential HSR Connector Service.)

    If you're operating lots of different train service types with vastly different average speeds you're doing something wrong and you're wasting a very large amount of other people's money. Don't Do That!

    Lastly, to claim that massive, years-long and corridor-long construction induced disruptions of Caltrain service due to CHSRA-promoted over-building are for the benefit of Caltrain passengers or other Peninsula residents is a shoddy fraud. The sole beneficiaries of promoting wasteful over-building between Redwood Junction and San Jose are those who will profit so spectacularly from BART Fremont-San Jose and HSR San Jose-Gilroy-Los Banos. This isn't for Caltrain or anything else; it's purely against Dumbarton/Altamont. Follow the money!

    3. Caltrain 5tph capacity "limit" isn't due to differing average speeds of Caltrain services, but to (a) a neanderthal-era signalling system; (b) particularly incompetent neanderthals placing and spacing the signals; and (c) the utterly insane regulatory and operating regime in which the non-performance and in-capacity of two irrelevant, insignificant daily freight trains sabotage and cripple the operation of 100 passenger trains. In short, Caltrain's signalling is designed to discourage coal trains in Nebraska from having too many collisions per annum; the difference in speed between Caltrain locals and expresses has almost nothing to do with it.

    4. You're going to hear a great deal more of this why-waste-money-on-the-Caltrain-corridor song, just as you have for the last 30 years; and you should expect to hear yeoman citizenry, disinterested elected officials, and tireless public servants from the whole region all spontaneously singing from the same VTA-harmonious songbook.

    It's not as if Quentin "BART to Millbrae" Kopp doesn't have a record of setting Caltrain service back by 20 years once already, and it's not as if Santa Clara...Brisbane won't spontaneously fall into line in exactly the same way that Fremont, Dublin, Livermore and San Francisco have already done whenever the occasion demanded.

    In the end, the most reasonable and most integrated and most sensible plan will emerge from MTC-facilitated "regional consensus", the simply awful, yawning gap in frequent electrified rail service between Santa Clara and Millbrae will finally be filled by BART, and San Franciscans will no longer need to endure a slow and noisy and bumpy and infrequent Caltrain ride to get to and from the regional HSR Megaterminal at San Jose Diridon Interdimensional.

  2. "The average speed of HSR trains on the Caltrain line will and in fact must be almost exactly the same as the average speed of Caltrain expresses, and Caltrain expresses aren't going to be averaging over 100mph.

    That's how it works everywhere else in the world, and that (minimizing the number of different traffic classes, maximizing intensity of traffic per track, minimizing capital and maintenance expense) is the only way it can work on the Caltrain corridor. (And it is and will remain the Caltrain (= S-Bahn/RER/Overground) corridor: 80% Caltrain, 18% HSR, <2% freight/dinosaurs. Tails don't wag dogs.)

    The minimum number of traffic classes is one local traffic class ... that's what the highest capacity, highest frequency metro lines operate. But around the world, interurban surface rail lines have more than the minimum, and support both local and express services ... in core metro areas, on quadruplicate track.

    Its quite odd to refer to "how it works around the world" in a screed arguing against a local/express set of four tracks between population centers the size of San Francisco and San Jose.

    Odd enough, perhaps, to require a vast amount of verbiage to distract from the core thesis.

    "Caltrain 5tph capacity "limit" isn't due to differing average speeds of Caltrain services, but ..."

    ... and then a list of things which can all be changed and still will not allow the mixing of genuine world class urban Express services and high frequency locals on a single pair of tracks.

    "The average speed of HSR trains on the Caltrain line will and in fact must be almost exactly the same as the average speed of Caltrain expresses, ..."

    ... if this conclusion follows from its set-up, then it also follows that "the average speed of Caltrain expresses will, and in fact must be almost exactly the same as the average speed of locals, ..."

    ... yet, at the sacrifice of track capacity, expresses achieve 60%+ higher speed than expresses. Which of course points to expresses able to operate at express speed and average 80mph mixing with rapid rail at a rapid rail speed of 110mph or (given the freedom from the FRA impositions on lines running those freight trains and take trucks off Bay area roads) 125mph.

  3. This comment has been removed by the author.

  4. BruceMcF: Its quite odd to refer to "how it works around the world" in a screed arguing against a local/express set of four tracks between population centers the size of San Francisco and San Jose.

    Dear Bruce,

    The whole point of professional, quality engineering is to solve problems in an efficient manner.
    It is not to start out with a slogan ("Altamont bad!" "Four tracks good!" "San Francisco bad!") or a particular solution ("Four tracks good!" "150mph good!" "Transfers bad!" "Freight trains good!") and then blindly burn money implementing it without regard of alternatives or opportunity costs.

    The fact is that a "world class" (how's that for a meaningless slogan?) rail transporation operation can be implemented and sustained in the Caltrain corridor using a modest amount of well-targeted infrastructure investment, developed in coordination with a service plan and in coordination with vehicle design and procurement.

    The fact is that to operate eight regional trains per hour per direction between San Jose and San Francisco (four local, four express, which is more than BART's busiest branch, with far better service quality and higher speed, and all of which could potentially be even higher capacity doubled-decked, double-length 300m+ length if it ever came to that) only requires and justifies a very limited amount of new track, which of course is expensive and disruptive to build.

    If you want to come and out say "I want to build four tracks everywhere because I read about it in a book or because that's how lines on the New York City Subway work and I dream than Caltrain needs to run 48tph" that's fine, but don't pretend that this project has anything to do with optimal solution of a transportation and economic problem in the corridor of the SF Peninsula.

    Consider also that these people, who have forgotten more about "world class" rail operations than US Professional Transportation Engineers will ever learn, come out and say outrageous things like: "For obvious financial reasons however, this [four-track lines] is not a generic solution to the problem, especially in dense urban areas, where underground building of a "second" line would have skyrocketed the costs. Moreover, full length doubling of a line is hardly justified economically. Full length 4-tracks lines are still a niche solution, intended to solve specific requirements."

    Corridor-long quadruplication would be nice in some abstract sense, but I'd personally rather spend the few hundred million saved dollars (and huge political capital) other than on a gold plating the tracks of my choo choo set, thank you.

    BTW in all the sloganeering about the population of San Jose, nobody seems to bother to either visit the place or look at a map: the city is geographically enormous, and much of it is nowhere near any existing or proposed Caltrain or HSR corridor. This low-density "population center" with a gutted-out, retail-free, largely employment-free, empty-lot-rich "downtown" is one that is going to be doing nearly all its moving in automobiles until the end of time. Sure, there should be an abundance HSR service and Caltrain service to San Jose, as much as the market will bear, but the number of riders will always be under a tenth of those from a real urbanized area with comparable population, while our local costs of Professional Engineering Contracting and construction are several times those in less corrupt societies. (You know: low-wage, low-cost, low-environmental-concern, low-construction-impact, low-land-cost places like the cities of central Europe.)

  5. Richard writes, The average speed of HSR trains on the Caltrain line will and in fact must be almost exactly the same as the average speed of Caltrain expresses...That's how it works everywhere else in the world.

    My world includes the NEC. Here are some sample cardings from the NEC:

    From Washington Union to Baltimore Penn (triple track):

    Acela Express 2182 (5 pm departure): 30 mins (78 mph)
    MARC 450 (express, 5:10 pm departure): 41 mins (57 mph)
    MARC 534 (express, 5:20 pm departure): 42 mins (56 mph)

    From Philadelphia 30th to Trenton (quadruple track):
    Acela Express 2172 (5:32 pm departure, does not stop in Trenton): 25 mins (79 mph)
    SEPTA 9747 (express, 5:15 pm departure): 51 mins (39 mph)
    SEPTA 4749 (express, 5:40 pm departure): 48 mins (41 mph)

    From Trenton to Newark Penn (quadruple track):
    Acela Express 2150 (6:58 am departure, does not stop in Trenton, but used to): 30 mins (96 mph)
    NJT 3922 (express, 6:58 am departure): 51 mins (56 mph)
    NJT 3924 (express, 7:13 am departure): 51 mins (56 mph)

    All along the NEC, the Amtrak expresses are running between 37 to 103 percent faster than the commuter rail express trains departing around the same time. I would characterize those deviations as being very far from "almost exactly the same average speed."

  6. Clem, not sure I followed your first point. - when you say total trip time from SF to LA "would increase" - would increase due to what? It seems like what you are saying is that at this HSR speed (125mph max, 94mph avg,) in the Peninsula, (is this a given?), HSR isn't worth it because it slows the entire HSR trip from SF to LA too much. Is this what you were saying??

    Secondly you seem to be pointing out that HSR could be an attractive replacement for Caltrain at 2X the speed... But to be a viable replacement for much Caltrain business, wouldn't it need to make many more stops for commuter purposes? And more stops just lengthens avg speed...
    (assuming riders would stomach 3X the ticket cost in any recurring fashion, for a few minutes of net time savings each day).

    So, when you calculated the 125top speed/94 avg, how many stops did you include? Zero? ( Then its not an apples to apples comparison with Caltrain...)

    Either you need to clarify your point here, or it sounds like you're arguing that the whole proposition of HSR down the Peninsula doesn't makes sense.

  7. Richard,

    Your posts aren't making sense. On the one hand, you claim that two tracks is clearly sufficient for Caltrain's needs. On the other hand, you claim that even with four tracks, HSR cannot run any faster than Caltrain. But if Caltrain only needs two tracks, then it's trivially obvious that with four tracks, HSR can run substantially faster with Caltrain (Caltrain only needs to use two anyway!).

    You're trying to have it both ways. You can either argue that four tracks is way too much infrastructure, or you can argue that even with four tracks HSR will not be able to operate faster than Caltrain. But you can't (credibly) argue both simultaneously.

  8. Anonymous -

    30 minutes implies 0 stops. Caltrain as it is currently built (basically 2 tracks the whole way) cannot support intermixing 0-stop express trains with 20-stop local trains. Can't do it at any reasonable frequency. It won't work. So it makes no sense to compare a 0-stop HSR train to a 0-stop Caltrain, because the latter is impossible on current Caltrain infrastructure (except late at night when there are almost no trains running).

    Clem is just pointing out that if you force people to switch to Caltrain from SJ to SF (i.e. eliminate HSR from SJ to SF), the total trip time from LA to SF will increase considerably. This will lower ridership. Similarly, if I forced you to walk from Palo Alto to SFO to catch your flight to Seattle, you would be less likely to fly and more likely to just drive all the way to Seattle or perhaps not go at all.

    Nowhere does Clem suggest that HSR would completely replace the Caltrain Baby Bullet service. He simply notes that the two new express tracks could benefit the Baby Bullets by giving them a lot of flexibility as to where/when they could pass the Caltrain locals.

    I have no idea where you would get the idea that Clem is arguing that HSR down the Peninsula doesn't make sense. You are completely misreading his post.

  9. @ Richard,

    I certainly agree with your description of the minimal improvements required to run more effective Caltrain service.

    But we're not just talking about Caltrain improvements; we're talking about adding HSR to the corridor. Why? Because "world class" requirements-driven engineering means that HSR (a) shall serve downtown San Francisco, and (b) shall link SF - LA in 2h38, or as close to it as reasonably feasible.

    Every second counts, which is why I think your suggestion to slow HSR to Caltrain speeds doesn't make any sense. HSR must reach SF as quickly as possible, and yes, hopefully do so without ruining Caltrain. I share your concerns about that.

    I think there is a clear derived requirement for four tracks because you HAVE to run traffic with different speeds and you HAVE to decouple the timetables of HSR and Caltrain.

    The need to accommodate HSR exists regardless of the Pacheco / Altamont clusterf***, and regardless of San Jose's pan-galactic aspirations.

    (I was just looking at the tangle of curves before and after the San Jose station: that will be solid 50 mph territory... gotta love that brilliant Pacheco choice)

    @ anon

    I'll have to clarify those points. I was talking about the scenario where HSR terminates in SJ and everyone has to ride Caltrain to SF (which is the "we already have a bullet" argument taken to its logical if unstated conclusion)

    94 mph is non-stop. Presumably most folks riding HSR north of SJ are going to SF, which has an enormous catchment area (east bay BART basin, marin, other northern reaches) ignored by Caltrain-centric thinking.

  10. Clem, let me begin by saying that I am very impressed by the level of
    “technical” and general rail-related knowledge displayed not only by you, but also by the several people who have added comments to your blog. Rafael comes to mind, but there are others who speak with authority and serious understanding. Indeed, these blogs seem to have avoided the ad Hominem road-rage rants that characterize the high-speed rail discussions on other blogs such as in the mid-peninsula online papers.

    And, this is what puzzles me so. How can there be such an unequivocal endorsement of a project by you and Rafael for which there are so many concerns and questionable issues?

    Is the personal and political history of Morshed, Kopp and Diridon of no concern to you as they promote this project? Several articles have been written that should certainly cast some doubt on their honesty and reliability. Even now, in violation of the law that the voters chose to support, they pursue their agenda without accountability or oversight.

    Then, there is the question about all the data and numbers provided by the CHSRA in all their Powerpoints, brochures and other modes of representation. Do you accept as certain fact all their environmental numbers, their comparisons, their costs, revenues and profits, the elapsed travel times, ticket/fare costs, ridership projections, cars removed, etc. etc.? Doesn’t all their PR rhetoric sound like an infomercial product that is too good to be true?

    Being a supporter of multi- and inter-modalities in public mass transit, both intra, regional and inter-city, I am certainly not in opposition to high-speed trains. Therefore, I endorse such trains where they are needed most; the two population regions in California where an integrated and adequate public mass transit is still missing. I can conceive of fast trains in the Bay Area, up to Sacramento, both sides of the Bay with ample connectivity with other modalities. Similarly in the LA Basin and Inland Empire. But, that is not what is on the table here. Indeed, Diridon took pains to belittle and humiliate me in public as part of his presentation strategy, when I asked him if this should not be considered.

    I am not equipped to discuss the number of tracks on the Peninsula but am willing to accept four as a minimum. I certainly am unqualified to enter the discussions that this particular thread is about. However, I do think that it is plausible to consider the very long, strategic view of the Peninsula in 75 years, what that will be like and what the expansion of rail service ought to look like. We are talking about the expenditure of many billions of what will surely be taxpayer dollars. Shouldn’t they be spent for the “long haul?”

    If the Peninsula becomes a metropolis of millions, as many now project, shouldn’t these expanded rail services, especially at such high speeds, be separated from the rest of high-density, high-rise urban development? Shouldn’t they be taken off the streets? That is, shouldn’t they all be put underground, as they are in most other US cities such as NYC, Boston or Washington?

    I know, the costs are extremely high. However, looked at within a time-frame of at least a century, wouldn’t these costs amortize over those years so that our grandchildren and their grandchildren don’t look back at our rail industries and ask if they were completely out of their minds to build them on 15’ high, 100’ wide retained-fill walls which divide the Peninsula like a spatchcocked chicken?

    I guess what disappoints me the most is the lack of a serious critical review of this enterprise, as distinct from the unchallenging endorsement of something that at least some have devoted a lot of time and energy to questioning.

    Forgive this being so long, but I have had trouble obtaining rational responses to my concerns.

  11. Do it right the first time so my generation doesn't have to redo it.

  12. @ anon @ 14:20 -

    promoting HSR is not quite the same thing as idolizing Kopp and Diridon. In particular, choosing Pacheco over Altamont-via-Santa-Clara may prove an expensive mistake if the "HST/commuter overlay" eventually has to be built as well.

    Still, electric HSR is the only technology that can transport tens of millions of people every year over intercity distances without consuming a drop of oil and, to do so safely, economically and with a small land footprint. The Iraq war, the ballooning Pentagon budget and the foreclosure crisis are just the most recent reminders of the direct and indirect costs of basing an entire civilization on the assumption that gasoline will always be cheaper than bottled water.

    In spite of substantial advances in battery chemistry and manufacturing techniques, electric cars in Joe Average's price range will not be drop-in replacements for conventionally powered cars for the foreseeable future. Rather, electric trains, electric cars and electric (folding) bicycles will complement one another.

    Of course, just because the idea of HSR makes eminent sense in California doesn't mean anyone should write a blank check for its implementation. For example, I've previously argued that the post of chairman of the CHSRA board (i.e. Quentin Kopp's job) should be turned into a directly elected office to improve accountability to the electorate. In addition, CHSRA should be kept on a short financial leash at all times, lest it permit project bloat or unnecessary gold-plated solutions.

    However, keep in mind that the whole concept of virtually dedicated HSR tracks is primarily a result of rigid, antiquated FRA rules on mixed traffic. If that agency allows Caltrain, UPRR and CHSRA to implement an integrated safety concept and timetable, then SF-SJ can become a rapid rail corridor (max speed 125mph) and make do with fewer than four tracks in some sections.

  13. There are underground mainline rail lines in Boston? That's news to me. The only reason that Penn Station is underground is because it's between two sets of river tunnels, while Grand Central's tracks are actually in a cut that was then decked over, though north of that, there's a much older tunnel under a hill, and north of 97th Street, the tracks are elevated. In DC, the tunnel takes the tracks under Capitol Hill. So unless you think that the whole Peninsula will have as much density as Midtown Manhattan combined with as many landmarks as Capitol Hill, your plan of building a 40 mile tunnel is absurd.

    and Rafael: to the best of my knowledge, Caltrain's explicit goal at this point is to get that FRA waiver. Oh and High Speed Rail isn't the only technology capable of transporting people over intercity distances without using oil. Plain ordinary Normal Speed Rail can do that too, albeit not as quickly, and in the LA-SF case, not quickly enough to be able to compete on speed with airplanes.

  14. @ arcady -

    precisely, normal speed rail could not attract sufficient ridership away from the airlines in the SF-LA corridor. Technology only matters if enough people actually use it to at least cover the operations overheads.

  15. @ anon 14:20

    Thanks for your comments. Perhaps I haven't made the purpose and scope of this blog sufficiently clear.

    (1) It is not about whether or not California's HSR should be built. While this question is often debated in other forums, I believe it is largely academic at this point in time. I assume (right or wrong) that HSR will be built.

    (2) It is not about whether to run HSR via Altamont or Pacheco. While Altamont is far superior and would have avoided a lot of peninsula impacts, the decision is political and has been made. I assume (right or wrong) that HSR will be built specifically between SJ and SF, along the entire length of the peninsula.

    (3) I will never hesitate to poke holes where the HSR plans don't make sense, discounting (1) and (2). I will endeavor not to "drink the Kool Aid".

    I believe in transparency and putting as much information out there so people can affect this project in a constructive and intelligent way.

  16. Anon, I don't think the Peninsula will ever become "a metropolis of millions." The Bay Area already has the lowest population growth rate of all US metro areas over 3 million people except Detroit. Its Central Valley exurbs are growing quickly, but San Francisco, East Bay, and the Peninsula have glacial population growth.

    Even if it were projected to have high population density in the future, there's no reason for tracks to be underground. Until the 1930s and 40s, most of New York City's rapid transit system was elevated, even in very high-density sections. In Brooklyn and Queens most of these els are still in place, many in neighborhoods that are middle class and denser than any part of San Francisco and the Peninsula.

    The Manhattan els were torn down in the 1940s, the idea being to replace them with subways. That program, the IND, was wasteful and overbuilt, and ensured the city would have no money to either expand or properly maintain the system. The take-home lesson isn't that cities should build subways now so that they don't have to later, but that building underground lines where at-grade or elevated ones will do can cripple rapid transit for decades.

  17. One interesting bit I picked up from the scoping meeting today: CAHSR and Caltrain will (almost certainly) operate together under a some sort of single entity. They are currently negotiating the specifics Memorandum of Understanding between the two groups. It will likely result in a new Joint Powers Board (JPB that includes a JPB??) or something.

    Though there will undoubtedly be politics involved in the specifics, but in general this is a good thing. This entity will probably be a be a single point for any negotiations with cities and responsible for the maintenance of the tracks, running the stations, and scheduling the 3 or 4 classes of traffic (Caltrain, Baby Bullet, HSR SF/Millbrae/PA/SJ, HSR SF/SJ express).

    Richard: as for you comparison to that Swiss paper, it's about a 40 mile stretch with 9 stops and 6tph. It has 4 classes of stops: 2, 4, 6, and 9. SF to SJ is 50 miles with 22 stops. It also will (likely) have 4 classes: 2, 4, 7, and 22. The lines are currently limited to 5tph, but with HSR this could easily hit a need for 15tph. There is simply no way to support 15tph, a 2-stop express going at speed, and a 22-stop local on only 2 or 3 tracks.

  18. @ Peter -

    creating a legal entity comprising both Caltrain and the peninsula section of the HSR network will also lead to a single dispatcher for train operations. This is essential for safety at high speeds and train count per hour.

    As for your Switzerland vs. the peninsula argument: how essential is it really to keep running trains that stop at each and every one of the 22 stations, especially during rush hour?

    I'm sure there'd be a hue and cry from anyone whose service level was reduced, but the Swiss example does suggest that something less than 4 tracks all the way through might be sufficient, provided that any given local train stopped at just ~50% of all stations (and the next at most of the others etc.) The timetable would be a bit of a puzzle for both operators and passengers, but the current "baby bullets" don't all stop at the same stations, either.

    Passengers traveling between secondary stops would have to transfer somewhere, but I suspect they are a small minority during rush hour.

  19. Doing contortions with scheduling timetables to get more capacity can make sense when the tracks exist and there's a strong incentive not to do any work. That's not the case here where there will be construction no matter what. If there's going to be major disruption of the corridor for construction the difference between 3 and 4 tracks is almost negligible. Doing a 3-track system would have almost the same construction detriment, wouldn't save too much money, and would cripple the long-term prospects of the system.

    As for the Swiss, I just looked up their timetable. Their express 2-stop is 33 min (72mph) and local 9-stop is 51 min (47mph). Our express 2-stop will be 30 min (100mph) and the local 22-stop will be about 75-80 min (40mph). So our fast is faster, our slow is slower, and were are likely to have more traffic. This is exactly the location that a "niche solution" (that's being built basically from scratch) is required.

  20. I don't see the need for a single entity, especially for any reasons of safety or even necessarily performance. Look at New Jersey for example. The NEC is owned and dispatched by Amtrak, and is used by both Amtrak and NJT trains. There are also various other NJT trains joining from other branches, which are owned and operated by NJT. It all comes together into one of the busiest two-track lines in the world, and most of the time, it seems to work just fine. Or look at Metrolink, which does the dispatching for its own lines and Coaster, and has Amtrak trains running over it, but Metrolink trains run on UP and BNSF lines, and vice versa, and it all more or less works too. Or heck, look at Switzerland, which has a mix of various cantonal and federal (and partially private) railways, with various companies using each other's infrastructure for their trains, and a coordinated dispatching system even between the Federal standard-gauge mainlines and the cantonal narrow-gauge branch lines, in order to provide guaranteed connections in case of delays. I just don't understand why it seems impossible for two bureaucracies to cooperate without creating a separate third bureaucracy for that purpose.

  21. It's actually quite common to have separate entities own / maintain the fixed right of way, versus operate trains. I imagine the expanded PCJPB could assume that infrastructure role, with private companies competing to provide service (subsidized or not) on the PCJPB's track.

  22. Clem, I can't find the poster's comment now, but someone said they asked about the San Bruno curve at the recent meeting, and they were told by the engineer they weren't going to fix it. That the slow down to 70mph (or smoething like that) was already built in to the operating assumptions. So, then... Lets say they're also not going to fix the Palo Alto curve issues..., or some of the other issues you've discussed by town. Are we still going to make it non-stop from SF to SJ in 30 minutes (which requires avg speed 100mph)

  23. I'll be taking a look at that. It's not rocket science.

    I have a post on curves coming up.

  24. Clem: right now, the PCJPB is just the bureaucracy managing the corridor and the service. The actual operations, maintenance of way, rolling stock maintenance, and dispatching are provided by their contractor, which is Amtrak. In Southern California, it's a similar situation, except the operator is Connex, and the maintenance of way and trains are done by two other private companies (I forget whom), while dispatching is done in-house, and actually makes a bit of profit from Amtrak, Coaster, UP, and BNSF trains.