10 February 2013

Keep Out the Coast Daylight

The Coast Daylight pulling into
Palo Alto in 1942.  Image from
Caltrans recently published an administrative draft of its California State Rail Plan. This document assembles hundreds of pages of assorted kitchen-sink gobbledygook, fails to mention important topics such as level boarding even once, and reveals a strange fixation with Amtrak's plans to start a new Coast Daylight train linking San Francisco with Los Angeles.

The Coast Daylight, as any old railroader will tell you, was one of the Southern Pacific's most prestigious services back in the golden age of steam trains. The red, black and orange livery of the massive GS-class locomotives (shown in the opening photo pulling into Palo Alto in 1942) is sufficient to throw even the most staid rail buff into convulsions of nostalgia, which seems the likeliest explanation for the sudden urge to resurrect this long-forgotten train.

Why should we care about this anachronism? As it turns out, this latter-day Coast Daylight would terminate at Fourth and King in San Francisco, causing a number of complications and constraints for modernizing the peninsula rail corridor.

Chronic Lateness.  The Coast Daylight's counterpart, the Coast Starlight, has the well-earned nickname "Coast Starlate."  Because the Daylight would also use hundreds of miles of track owned by freight railroads and subject to all sorts of delays, the northbound Daylight would be exceedingly unlikely to arrive reliably on time, causing it to miss its assigned timetable slot on the peninsula rail corridor and delaying everybody else.  With track capacity in the blended Caltrain / HSR system a scarce and valuable commodity, one must ask, should all passengers (especially those who value their time and use high-speed rail) have to pay for Amtrak's inability to keep to a timetable?

Diesels Forever.  The Coast Daylight would be a diesel train, and as the State Rail Plan notes, it could not use San Francisco's new underground Transbay Transit Center station where diesel exhaust is not allowed for.  This would strand it on the surface at the 4th and King station, which is increasingly becoming the object of San Francisco's desire for urban redevelopment.  Plans for Amtrak trains to San Francisco clearly clash with San Francisco's plans for the surface rail yard, a clash that wouldn't arise with 100% below-ground electric trains.

Yet Another Platform Interface.  The Coast Daylight would presumably use the same equipment as other Amtrak long-distance trains, with an entry floor height of 17.5 inches.  No matter what floor height Caltrain ultimately selects for the necessary upgrade to level boarding, Caltrain platforms will end up higher than this.  Because steps down from the platform into a train aren't allowed under ADA and FRA regulations, the result would be separate platform tracks entirely dedicated to the Coast Daylight at San Francisco, Millbrae and Redwood City--or no level boarding for Caltrain.  That hardly seems like optimal use of expensive station facilities.

Negative Return on Investment.  Thanks to speedy and frequent service, the lucrative San Francisco - Los Angeles travel market will go mostly to HSR, with only marginal ridership left to the Coast Daylight to pick up in coastal communities in between.  The Coast Daylight will join many other Amtrak long-distance trains with subsidies per passenger well above the price of a ticket.  The opportunity cost of every dollar spent on reviving the Coast Daylight means that rail service will languish in areas with far greater potential.

Yet Another Tenant Railroad.  Caltrain's plans for modernized train control (known as CBOSS) make a big deal of accommodating so-called "tenant railroads" that travel over Caltrain-owned tracks.  While the Coast Daylight has indeed been accounted for by the Diesel Brain Trust, the very real possibility that CBOSS might fail and get replaced with the HSR train control system could make integration of the blended system unnecessarily difficult.

Blending different services on shared and limited rail corridor infrastructure is a good idea in principle, but blending can go too far.  Amtrak is the spice that will make this blend go sour.  The Coast Daylight should terminate in San Jose or Emeryville, and even nostalgic rail buffs must accept that Amtrak should keep out of the peninsula rail corridor.


  1. It occurs to me that the Coast Daylight is also important precisely because it covers the coastal corridor, which would otherwise have no viable service. Yes, HSR would pick up the luxe LA-SF market, but I for one would be for a piloting to see how much latent market demand there is along the coast (e.g. in the Monterey-Salinas market).

    But I also agree that introducing it to the Peninsula corridor would be excessively onerous. What about running it up the East Bay to Jack London instead?

    1. You're conflating "covers the coastal corridor" (dubious in itself) with "important" (most assuredly not) and with "viable" (most certainly not).

    2. No, I'm simply pointing out that there's a latent market on the corridor. We'd need a pilot to see if it can be viably tapped.

  2. Huh. I never realized that the new Coast Daylight would only be LA-SF (and I never realized that was the old route either). I had been expecting a second long-distance train on the same route as Amtrak's Coast Starlight (Seattle-LA).

    Does anyone know where the delays usually crop up along the Starlight route? I suppose everywhere is the answer, but an LA-SF train should be considerably more reliable than an LA-Seattle train.

    1. The worst delays are up near the CA-Oregon border but none of it's particularly reliable, and UPRR has little interest in improving Amtrak service.

  3. How about Sac -> Oak -> SJ -> central coast -> LA -> SD? This would be somewhat useful service for the central coast. I'm sure it would be faster to run more buses on 101, but the bus-train transfer is worth an hour extra travel time to some people.
    This service would probably be expensive and require a higher subsidy compared to more buses, but it would be a better use of Caltrans gas tax money than a highway expansion along 101, and politically the Central Coast would be getting construction and transportation dollars, without yet another subsidized highway expansion.
    When SF-SJ and LA-SD are upgraded, the service could be shorted to San Jose<-> LA via the coast, with transfers to the Capitol Corridor/Caltrain/BART/HSR and the Surfliner/Metrolink/Metro/HSR at either end.

  4. Now, if Caltrans bought some good tilting Pendelino or Talgo trainsets, and got FRA permission to use them at full speed around the many curves along the coast, is it possible that the rail route along the coast could be competitive with the highway? I'm thinking about the modeling done at http://zierke.com/shasta_route/ for improvements to the existing Coast Startlight between Redding, CA and Eugene, OR, via Yreka/Ashland/Medford/GrantsPass. According to this table, proper tilting trains and regulations could improve speeds from a current 39 mph limit, up to 65 mph around a curve of 1000m radius, for example: http://zierke.com/shasta_route/pages/05curve-curve.html

    By the calculations on Zierke's website, the Coast Starlight from Sacramento to Dunsmuir (near Mt Shasta) could be sped up from 5:05 to 3:14 travel time, almost 2 hours time saved for 216 miles. With similar improvements due to tilting trains and better track maintenance, we could imagine improving San Jose to Santa Barbara from the current 8 hour schedule down to 5 hours, about 30 minutes slower than driving w/o traffic. If the Surfliner tracks are also improved, the total trip might be 7 hours from San Jose to LA, certainly too slow for the whole route, but similar to driving for someone coming from San Luis Obispo or Paso Robles, where I-5 isn't a competitive option.

    If California and Oregon were interested in subsidizing even more service to the hinterlands, they could continue the trains from Sacramento north thru Redding and Klamath falls to Eugene and Portland; it would still be a slow trip, but at least it would fall at a better time of day, and the tilting trains could save a few hours off of present 16 hour timetable.

    1. The problem is that UP is not going to allow Talgos to be used at full speed, let alone Pendolini. BNSF, a generally passenger rail-friendlier operator, restricts the Talgos on the Cascades to 5" cant deficiency, the same as a non-tilting Amfleet on the NEC.

  5. One more point to run a Daylight to Oakland- That's where the maintenance facility is. Trains need to be restocked- tanks flushed, cleaned. Those facilities don't exist in SF.

  6. As it turns out, the coast route isn't even a good way for conventional trains to get from the Bay Area to LA. Taking the San Joaquin to Bakersfield and then transferring to a bus saves you nearly 4 hours on Oakland-LA compared to the Starlight. If you're going to put in the money for additional service in the short-medium term, might as well spend it to finish doubling the Tehachapi line (could probably get a bunch of the money to do so from UP and BNSF too). Plus it would make the ICS marginally more useful, though running 32t/axel locomotives over new high speed track is probably a bad idea to begin with.

  7. Clem: you persist in assuming that CalTrain will have level boarding -- despite Caltrain having no plans for level boarding, now or in the future?

    Which do you think will happen sooner: Caltrain planning for level boarding, or Republicans acknowledging anthropogenic global climate change?

  8. Lets step back and relax a bit.

    1) What is the current status of the Coast Daylight? Not running.

    2) How much significant political effort is behind it?

    3) How much more likely is it to run to SF? In the grand scheme the Coast Daylight could really be the Coast Starlight offset by 12 hours?

    4) Why pick a battle with other train advocates?

    This whole non-problem becomes even less of a non-problem if the Coast Starlight becomes a twice a day train following the same route up to Seattle.

  9. Coast Daylight- agree keep it off Caltrain/HSR tracks- it will be a concession /tourist train mainly serving pax in intermediate cities (like the San Joaquins) and people with time on their hands- so a transfer at San Jose or Oakland/Emeryville is acceptable for those needing to get to SF/peninsula destinations. Since the Coast Line doesn't see many freights, timekeeping should be better than on the much longer Starlight.

  10. The problem with this whole discussion is that neither the Coast Daylight nor the HSR actually exist. They're both still at the purely imaginary stage, and while the Daylight is allegedly only a year or two out, it's been a year or two out for a while now. Still, if it actually does start running, it would be a good decade or two earlier than HSR, especially HSR that actualy reaches the Bay Area, and probably a couple years before Caltrain electrification (which itself has been 5 years away for as long as I've been in the Bay Area). There's nothing wrong with diverting the Daylight from SF to Oakland once it's no longer possible to run it on the Caltrain line. Also, I suppose I should point that the East Rail in Hong Kong runs a much more intensive service than Caltrain, yet still manages to mix in intercity service, including long-distance sleeper trains, and at one point even had freight service.

  11. I don't understand why Amtrak California NEEDS a "Coast Daylight". It has one. Just because the through train is called the "Coast STARlight" doesn't mean it's a night train for its entire trip. From Jack London to LAUPT the schedule is Depart 8:50 AM, Arrive 9:00 PM. For LAUPT to Jack London it's Depart 9:25 AM, Arrive 9:32 PM.

    As is pointed out above, the Coast Line has few freights; they're mostly pickup and delivery locals. There just isn't much of a market for Bay Area to LA rail freight; they're close enough via I-5 that trucks can make the run with one driver departing early in the morning and delivering before end of business the same day.

    So the northbound Starlight is pretty darn reliable; all it has to worry about are MetroLink and Surfliners which also run on a schedule. The southbound, not so much, though there is quite a bit of pad in the schedule south of Redding.

    The point is, this train would be redundant. And Amtrak California already tried an overnighter; that was a fiasco and ended in less than two years. This proposal should have already been killed; it's probably on the table simply because nobody with any authority has thought it serious enough to oppose it.

  12. In its current condition, the Coast Line is not well suited for Bay Area - LA trips. It is suited for intermediate trips. Maybe more SLO -> LA and Bay Area trips are the answer. More logically, Paso Robles -> Bay Area, as the trip over the Cuesta Grade is SSSSLLLLLLOOOOOWWWW.

    Do the studies or entities promoting the CoastDaylight know what the market is? As I understand the proposed schedule, it's like an hour or two off the existing Starlight schedule. What if trains from SLO went north and south to fulfill trips into SF and LA in the early morning? Who is the market for the Daylight?

  13. Frankly, I’m appalled at the lack of imagination on this topic. Is the route slow? Yes. Does it need upgrades? You bet. Does every route need to be a speeding bullet? Not at all! Some people actually want to be on a train for a good part of the day. The Daylight will provide an alternative to the Valley based HSR route. Another positive is that in the future, passenger service will return to the Santa Cruz Branch—producing a true passenger network around the Bay Area and the rest of Northern California.

    Some logical solutions:
    1. Daylight: #798 leaves SF at 0520, arrives in LA 1700; #799 dp LA 0905, ar SF 1935
    2. Starlight: #11 operates six hours later than its current schedule; #14 dp LA 1625, dp Emeryville 0407, ar SEA 0245
    3. New route would operate on Starlight’s current schedule north of SAC but south of the capital city, the trains would actually traverse the San Joaquin route and serve Valley cities enroute to LA
    4. Standalone Shasta route between SJ and POR would serve daytime passengers
    5. Track upgrades on the Coast Route would then make travel times faster
    6. Additional frequencies on the Daylight route means SF-LA passengers would have three roundtrips on the scenic route
    7. Engines: Either conventional routes like the Daylight would switch over to electric engines or they would terminate at Caltrain’s 4th Street Station

  14. So I get up at 4am to get to LA by 5pm? Wow. Just wow...

    1. Yeah, twelve hours is just too slow. Get it down to eight, and maybe with a one hour turnaround, you can utilize the trainset for the return trip, rather than a costly layover.

    2. It's already eight via the Central Valley. All you need to do is spend the money on double tracking Tehachapi rather than on whatever improvements UP would almost certainly require along the coast.

  15. Just have the Coast Daylight either terminate in SJ or OAK. There will be at least one FRA dinosaur platform track going through Diridon, just keep the damn thing off the main Caltrain corridor north of Diridon.

  16. I don't think these are unresolvable problems. The lateness problem should only be an issue with north bound trains arriving into Diridon late that got stuck behind hopper cars filled with tomatoes somewhere along the coast route between Gilroy and LA. Caltrain just needs to enforce it's schedule like airports manage take-off slots.

    Caltrain could designate time slots N.B. Coast Daylights would be allowed to depart Diridon (ex: 5:05, 6:05, 7:05 ect.). If the 5:05 train was late by any amount it would wait in the station until 6:05 to depart. Amtrak would then notify their passengers of the delay and inform them they can transfer to any of the Caltrain's that will be departing in the meantime. Amtrak could negotiate with Caltrain to accept Daylight tickets as Caltrain fare. Otherwise the passengers would be free to stay on the Daylight and accept the fact they may arrive exactly 1 hour late. Therefore regardless of how late the Daylight is it would need to adhere to allocated time slots in Caltrain's schedule.

    The less station stops the Daylight makes on the peninsula, the less issue with differing platform heights (assuming Caltrain changes to high platforms). I am guessing Palo Alto might be the one and only peninsula stop for the Daylight. If Palo Alto had high platforms couldn't they build secondary low platforms along the same tracks just south of the primary platforms (adjacent to the station parking on both sides of the tracks)? Extend the platforms a bit further south and they could also make use of the pedestrian under-crossing at the end of Homer. There would be no problems accommodating a few low level platforms at Diridon and in SF the Daylight could terminate at 4th & King which could keep a few low level platforms as well.

  17. Here's how the pacific coast rail corridor would look like if improvements were made (i.e. electrification)

    https://maps.google.com/maps/ms?msid=202278891749402454243.0004d12e3f5eee0b54b45&msa=0 (Chatsworth to San Diego)
    (Chatsworth to SFO)
    https://maps.google.com/maps/ms?msid=202278891749402454243.0004cfb056b5034a7b9b2&msa=0 (Bay Area)

    Also to mention, how this bi-level railcar. (links: http://www.modelena.ru/tiporazmer_HO/podvignoy_sostav/bachmann_2011/passvag/e202.jpg, http://www.yaplakal.com/pics/pics_original/7/0/0/274007.jpg, http://www.yaplakal.com/forum2/topic282799.html, http://trs-msts.ru/index.php?do=photo&a=showphoto&photoid=175; http://kvisaz.ru/files//2011/01/15-bilevel-rail-car-russian-61-4465.jpg (new bi-level), http://www.hebners.net/amtrak/amtSUPER/amt32501.jpg (Superliner)) This design is from Russia and looks similar and functions identically to the Superliner series and has high level platforming. The only changes that are made are to re-gauge it to standard gauge.

    1. Russian bi-levels are the same overweighted s**t as American one. But in USA that was caused by an idea of "safe" train, while in Russia it was made to allow passenger trains to be used in the middle of freight trains (e.g. trains for military and rescue personnel+machinery)

    2. passenger cars to be used in the middle of freight trains*

    3. How come UPRR is so stingy with their tracks, even the less-used lines like the Los Angeles-San Jose track that the existing Surfliner and Coast Starlight use?

    4. Because they have no financial incentive to do otherwise. Getting paid for a few FRA-compliant trains per day to run makes sense, but once you get higher frequencies, or try to run non-compliant trains, you begin seriously disrupting freight traffic.

    5. really? I mean the freight traffic within that ROW is infrequent compared to their line that serves the Central Valley

    6. You're assuming that UPRR's business profit maximization motive is always in running freight trains.

      Think about it!

      Fact: the "public" (meaning contractor puppet) agencies have made one thing unambiguously, crystal clear over the decades, and without any possible question since Altamont was killed, and that thing is "cost is no object".

      "Holding out for more money" is, without a doubt, the correct and guaranteed to pay off business strategy for a company with real state holdings, just as "collude to drive costs through the roof" is the correct strategy for local rail engineering, civil engineering and construction companies.

      UP is being perfectly rational and acting in its shareholder interests.
      CSHRA and PCJPB are also acting perfectly on behalf of private interests.
      Everybody is a winner!