16 April 2011

Phased Implementation

UPDATE 4/30: A more detailed memo has now been posted for approval by the CHSRA board this Thursday. It describes the first phase, the last phase (full four-track build-out), but nothing in between describing what all these intermediate phases might actually look like.

Original Post: The California High-Speed Rail Authority has now posted their description of what 'phased implementation' means for the peninsula. The Phased Implementation Fact Sheet and Phased Implementation Q&A are the first documents posted to the CHSRA's San Francisco - San Jose document library following an eight-month dry spell. Look for this to become the new buzzword.

35 comments:

  1. Am I reading these too simplistically if I take them as saying that HSR will share Caltrain tracks until 2035?

    ReplyDelete
  2. Adirondacker1280016 April, 2011 13:56

    Ideally HSR and Caltrain would share tracks until they build a vactube system between San Francisco and Los Angeles or teletransport in perfected.

    ReplyDelete
  3. I guess the logical question is, is there any benefit in electrifying before grade separating? Will the expense in tearing down and rebuilding electrification infrastructure during grade separation outweigh the benefit of getting HSR service on the peninsula before grade separation is completed?

    ReplyDelete
  4. @Jon: constructing a grade separation on an electrified railroad is no big deal ... happens all the time. So no, that's not "the" logical question at all.

    ReplyDelete
  5. Adirondacker1280018 April, 2011 17:27

    Grade separating electrified railroads happens frequently. Does it make sense to install electrification in 2017 and then come back and rip it all out in 2022 to do the grade separation?

    ReplyDelete
  6. Catenary = cheap. Substations = expensive but will still be around.

    ReplyDelete
  7. Adirondacker1280018 April, 2011 17:59

    No catenary even cheaper than put in catenary and then rip it out a few years later. Might even be called "free"

    ReplyDelete
  8. Extra diesel locomotives are expensive. Best industry practice is to electrify connecting lines, even when they'll be bypassed or replaced in the near future. The Koreans didn't stick a diesel locomotive in front of the KTX south of Daegu; they electrified the legacy line and ran unmodified trains on it at lower speed.

    ReplyDelete
  9. Adirondacker1280018 April, 2011 19:12

    Electrify the main line in Bergen County NJ in exchange for the excess diesels and Comets. Or the BNSF in Chicago.

    ReplyDelete
  10. I don't see the problem with building grade separations after electrification. After all, the substatons will still be there and needed, as will all the wires, poles, etc. with fairly minor changes. And don't forget all the other things that go into the cost of electrification: design work, signal immunisation, grounding and bonding, and of course rolling stock.

    ReplyDelete
  11. You have to consider that Caltrain is planning to pay over $100,000 for each and every catenary mast (including labor and materials), so it might be pricier than first thought.

    Then again the stuff goes together like an erector set, so you'd be out only the labor cost (probably about $95,000), not the material cost (probably about $5,000).

    ReplyDelete
  12. Adirondacker1280018 April, 2011 20:04

    fairly minor changes.

    Lets assume they make the careful design decisions to place the catenary supports in exactly the place they will be after grade separation. Keeping the numbers nice and round, so that the catenary is 20 feet above top of rail. They then come along and raise the rail 6 feet. The arcing between the catenary and the train body will be a minor problem.... The way the pantographs get tangled in the nice fragile insulators will stop it. When the catenary falls down... All very minor.

    ReplyDelete
  13. You have to consider that Caltrain is planning to pay over $100,000 for each and every catenary mast (including labor and materials), so it might be pricier than first thought.

    Let's say 10 masts get moved for a grade-sep (total: $1 million). A major grade sep project can easily be 10-50 times that amount, so in the grand scheme of things not something to worry about.

    On the other hand: staging a grade-sep on live rail line carrying premium HSR service, that can get complicated -- esp. for an outfit that couldn't even manage a trivial BB upgrade without a complete weekend shutdown.

    ReplyDelete
  14. Adirondacker1280019 April, 2011 00:04

    More like 125K to move a mast. Just because you take down the old mast doesn't mean the bond issued to pay for it evaporates. It costs money to take it down and re-erecting it is going to cost almost as much erecting it.
    There's going to be quite a few changes along the Caltrain ROW. Nice round numbers 1,000 masts at 100k a pop is 100 million. Ya can buy a lot of locomotives and cars, that have significant resale value, for 100 million, Run diesels until the final configuration is in place.

    ReplyDelete
  15. "You have to consider that Caltrain is planning to pay over $100,000 for each and every catenary mast ..."

    Fun fact to know (and slits your wrists over): Caltrain's genius World Class expert engineering consultants are specifying 50 yard spacing for catenary stanchions on straight track.

    In contrast, actual engineers who have the slightest clue about what they're doing, such as for example those at DB Netz who developed the national standard "Re200" catenary specification (fully integrated with track planning CAD of course) allows 80m spacing for V≤200kmh, R≥1750m and wind ≤25ms-1.

    So, 75% more poles, with more clutter and more hardware and more visual impact and more maintenance cost ... and the all important, overriding, more capital (mmmmm... pork...) expense. Kalifornia über Alles. Synergistic! World Class! Globally unique! We'll show those Yurpeens how it's really done. Bigger and fatter = more American. Winning hearts and minds.

    The charitable interpretation is that Caltrain's World Class engineering consultants got something horrible wrong when they, at huge public expense, translated the units that every single piece of equipment will be manufactured in into the 17th century furlongs and slugs that are all they know.

    The more realistic and far more likely interpretation is that they are utterly ignorant, professionally incompetent, and unfit to be in charge of even an HO scale model train set in their own bedroom.

    ReplyDelete
  16. @Reality Check- it is a logical question, as shown by the discussion above. Maybe think before responding, especially if you can't manage to be polite?

    ReplyDelete
  17. "I guess the logical question is, is there any benefit in electrifying before grade separating?"

    Once upon a time I would have said yes, because once upon a time I believed that the political facts on the ground of the electrification fixed infrastructure, combined with a radically improved accompanying train fleet and radically improved operating efficiencies (from escaping FRA regulation, not from a change of power source), was the key step that needed to be taken to establish and grow high quality and modern transit on the SF Peninsula.

    But Caltrain is, of course, planning none of that. They just want to spend money,
    and the more stupidly and less strategically and more subject to do-overs the better for them and their contractors.

    (Today I'd probably put doing a four-track timed-transfer Hillsdale grade separation and station project as the second highest priority, after the first which is to liquidate Caltrain and everybody who works or contracts there. Electrification next.)

    "Will the expense in tearing down and rebuilding electrification infrastructure during grade separation outweigh the benefit of getting HSR service on the peninsula before grade separation is completed?"

    First, the "benefit" of HSR service on the peninsula is entirely negative. A Flight Level Zero airline in your backyard is not anything that anybody should have to put up with, let alone be propagandised at and mislead and misinterpreted about and misportrayed as a positive.

    Second, doing things once is always cheaper than doing them twice. But the reality for Caltrain is that their agency "engineering" overheads are so outrageous and their simple civil construction costs are so astronomical that the couple hundred or so extra man-hours of construction to erect new overhead, cutover the old, and test the new is almost certainly way way way down there in the noise. Add to that they won't be (shouldn't be, but God knows) moving substations or interlockings and the like, but just doing the comparatively trivial overhead knitting, and one could make another argument that the overhead's overhead ought to be neither here nor there.

    Against that is the fact that I'm sure we will be burdened by special Only In America safety regulations that make construction anywhere within a half mile of a live 25kv line effectively impossible. (Compare to this for example.)

    ReplyDelete
  18. Adirondacker1280019 April, 2011 14:04

    A Flight Level Zero airline in your backyard

    They aren't proposing a Flight Level Zero airline on the Peninsula. They are proposing a modern version of the current NEC or Keystone Corridor. It would risk turning Palo Alto into the slums of Princeton. Or something Metra Electric like. Wouldn't want to risk turning Atherton into Hyde Park would we?

    ReplyDelete
  19. They are proposing a modern version of the current NEC or Keystone Corridor.

    They are proposing an entirely separate system between Santa Clara (CP Coast) and Brisbane, with a separate signaling system. Does that look like your experience of the NEC or Keystone Corridor?

    ReplyDelete
  20. Adirondacker1280019 April, 2011 21:45

    They are proposing an entirely separate system between Santa Clara (CP Coast) and Brisbane, with a separate signaling system.

    Amtrak and the various commuter agencies and freight companies maybe be incompetent but they aren't stupid. They either use the legacy system, where it still exists or ACSES.

    ReplyDelete
  21. Clem,

    Are they presently proposing a sealed HSR corridor from Santa Clara to Brisbane, with this, or is that rather what they previously proposed under prior management, since departed?

    ReplyDelete
  22. Dedicated HSR tracks have always been the end goal.

    ReplyDelete
  23. "have always been" is an easy way to overlook a change at the top.

    Dedicated express tracks each way is a functional requirement of funding the Caltrain corridor HSR improvements with Prop1A(2008) bonds.

    That implies that Caltrain has:
    Local tracks = Total tracks - 2

    Dedicated HSR tracks as a CHSRA planning position is easily understood as a way to avoid having to think through problems. However, they have to think through those problems in any event with phased implmentation.

    ReplyDelete
  24. a way to avoid having to think through problems

    And you wonder why some of us don't trust them?

    ReplyDelete
  25. The phasing "problem" was solved a long time ago. It's simple and obvious and takes no great skill or intelligence to develop or implement.

    http://www.pobox.com/users/mly/Caltrain-Timetabling/201104-takts.pdf

    (Also blindingly obvious: fully twice as much service supported using the same corridor infrastructure if HS isn't forced to detour via Los Banos, providing worse service at $20+ billion of incremental regional transportation system expense.)

    ReplyDelete
  26. Actually, might it be possible to shift things such that HSR is coordinated with CalTrain locals at Redwood City (given a little bit of additional infrastructure, of course)?

    ReplyDelete
  27. Joey, a timed- overtake-transfer station really wants quadruple track to extend one stopping station upstream.

    The keen-eyed observer will note that I don't have this in the not-rocket-science diagram (sorry it didn't come out as a link the first time) to the north of Hillsdale.

    That's because Hayward Park is gone (as it will) and because I'm allowing triple-tracking though downtown San Mateo to temporarily slide into the "too hard" basket.

    The analogous situation for Redwood City would require track amplification south of Menlo Park. (Atherton gone, as it will be.)

    There's a penalty for trying to do real-world compromise of course, and I honestly show it: southbound local trains into Hillsdale wait an extra two minutes at the station to allow the express to catch up, a small operating efficiency traded against a much smaller initial implementation cost.

    (The same delay is shown for NB locals, though in practice such trains can follow a bit closer since the leading train's departure time is coordinated. And just FYI this is what BART does SB/EB at MacArthur a zillion times a day, with Fremont trains departing behind SF trains on a single track through downtown Oakland.)

    After some years, as part of a coordinated and prioritized corridor improvement plan, in some sort of alternate reality world in which Caltrain actually implemented capital projects that provided service and in which Caltrain implemented grade separations and station reconstructions in a sensitive and attractive fashion, the money and political capital would eventually come to relieve this bottleneck through downtown San Mateo, thus speeding up local trains and, more importantly, improving the robustness and flexibility of corridor operations in the presence of inevitable disruptions.

    But a RWC overtake (which I half-way suspect might be a marginally better location from a passenger perspective, though it isn't certain) requires a much longer and otherwise-useless build out. It also interferes with the alternate reality world nice burrowing junction just south of RWC where the high speed line departs the Caltrain line and trenches its way out under the old Dumbarton ROW and under the Bay. And politically as well as logistically (the latter is primary) there's a great deal to be said for not over-building through PAMPA or anywhere south.

    On the whole I judge the construction and cost advantages of Hillsdale to outweigh whatever urban centrality advantages Redwood CIty might be argued to have. It's also easier to make space for a truly bitchin' bus/train interchange at the quasi-greenfield Hillsdale site.

    The great thing about Altamont is that a very limited amount of extra track is required, that it is all confined between Redwood City and Bayshore, and that Caltrain can operate a truly attractive and effective and compromise-free schedule, forever, on a just pair of tracks south of Redwood Junction and north of Santa Clara.

    ReplyDelete
  28. Adirondacker1280030 April, 2011 14:45

    The phasing "problem" was solved a long time ago.

    Where do the express from Sacramento and the local from Stockton fit in? The hourly service to Los Vegas? What happens on Thanksgiving weekend when they want to run 15 trains in the peak hour(s)? Everything runs at the speed of the locals for 20 years while they complete the EIS etc. for expansion to four tracks for the areas between Bayshore and Redwood City that aren't four tracked?

    ReplyDelete
  29. As anybody who has been reading for a while knows, "Adirondacker12800"'s questions aren't serious and don't solicit information, so answering them with an audience solely in the Adirondacks would be wasted effort. But perhaps for others:

    "Where do the express from Sacramento and the local from Stockton fit in?"

    Under Los Banos HSR there is no feasible Sacramento-Bay Area service.

    Under Altamont it terminates in San José (the Capital of Silicon Valley, you'll recall), with timed transfers to any of the four (or more!) available Dumbarton-SF slots.

    Another alternative is cross-platform transfer to Caltrain express in Redwood City. (This is actually quite a nice service pattern, with a 15 minute turnback time for the Altamont shuttle train.)

    Or you use the four extra available (but not explicitly diagrammed, because they're not real world relevant) HS slots per direction per hour to run to SF, where trains all pile up in a big heap, to somehow be dealt with by the Tooth Fairy.

    "The hourly service to Los Vegas?"

    Not a serious question.

    "What happens on Thanksgiving weekend when they want to run 15 trains in the peak hour(s)?"

    What happens when the sea level rises and downtown San Francisco is abandoned? What happens when robots turn on their masters? Who would win if Santa Claus got into a fight with Godzilla?

    There are actually 8 HS Altamont-Transbay slots available every hour, but this leads to extreme congestion at Transbay, which anybody who has paid the slightest attention any time in the last 15 years knows is the limitation on line capacity.

    There are also several other limited-stop (but not identical with the illustrated express pattern) slots available each hour, which could, in some completely hypothetical alternate universe in which eight trains per direction per hour of Caltrain together with 10 lanes of Hwy 101 + Hwy 280 were inadequate to local mobility needs, be used by peak-of-peak extras ... but again putting pressure at the terminal where it can least be borne.

    ReplyDelete
  30. I was really talking about Altamont, where having a HST and CalTrain local both enter Redwood City from the south at the same time would be more or less trivial.

    ReplyDelete
  31. Adirondacker1280030 April, 2011 21:54

    Under Los Banos HSR there is no feasible Sacramento-Bay Area service.

    There isn't under Altamont either. The line to Sacramento doesn't get built until phase 2. Nice phased implementation of Caltrain upgrades puts off Redwood City to San Jose until they are "needed" The NIMBYs, the upgrade-it-when-it-needs-it masochists and the we-are-broke Teabaggers manage to keep diesels running to San Jose until 2075.
    Anyway BART to Fremont is faster.... two can play this game.

    Under Altamont it terminates in San José (the Capital of Silicon Valley, you'll recall), with timed transfers to any of the four (or more!) available Dumbarton-SF slots.

    In Fremont?

    Another alternative is cross-platform transfer to Caltrain express in Redwood City.

    The wye from Dumbarton is going to hovering over the Redwood City station? Or will the trains to SF go through Redwood City and the trains to SJ head south a bit south of the platforms in Redwood City? Where the wye to SF or SJ is? I suppose they could carve out a second ROW from the Bay crossing so that the trains heading to SF are pointed north through Redwood City and the trains headed to SJ are pointed south. Sounds expensive. So does the hovering over the station option.

    or "you mean cross platform transfers to Caltrain in Fremont." Caltrain is going to have service to Fremont?

    ,where trains all pile up in a big heap, to somehow be dealt with by the Tooth Fairy.

    No need to do that. Amtrak manages to get Acelas in and out of Penn Station in 15 minutes. NJTransit and Metro North do it in 10. Rumor has it that NJTransit and Amtrak manage to do it in 4 on some non revenue moves. I'm sure they could spare a few dispatchers and BLE members for a week or two to show them how to do it fast enough to get trains in and out of a station in reasonable times.

    "The hourly service to Las Vegas?"

    Not a serious question.


    I come up with 3:30 for a super-express. Non-stop flight is 1:30. With arrival at the airport an hour ahead of the flight for 2:30. Personally I don't cut it that close but lets call it an hour. Airplane seat to curb is a half hour under optimistic conditions. 3:00. Then there's incentive of going through security and the wonderful seat arrangement on most planes. And the fine dining on hour and half flights. I guess it's not a serious question. Or someday the trains to Reno.

    Who would win if Santa Claus got into a fight with Godzilla?

    Santa Claus. On the other hand no long distance train operator ever has to run extra trains when demand is high.

    http://www.youtube.com/watch?v=jXbs1uq-NJU

    Don't take the Amtrak trains that are numbered 3xxx or if you are traveling between NY and DC the ones that originate in Boston or Springfield.

    ReplyDelete
  32. Joey: re Redwood City cross-platform transfer for Altamont trains: Yes, you're right. Sorry to have misunderstood what you were hinting at.

    If you look at the (unassailably correct and only rational choice!) "15/15/15A" build-out scenario in the diagram, you can see where there's a very clear and nice opportunity at xx:18 (and xx:48, and xx:03 and xx:33 if you go that far) for a train to come in off the Dumbarton line, stop and terminate at the northbound local (inner) track just as the northbound Caltrain express is arriving across the platform on the northbound express (outer) track. Then there's time to run through a crossover to the other local track (next SB local not due until xx:35), reverse, and wait on the SB local track for the xx:30 connecting arrival of a SB Caltrain express, and then depart back across Dumbarton.

    Tick-tock clockwork! Warms the heart. It almost resembles a real train network, with the half-hourly (or quarter-hourly, in some far off future) expresses connecting cross-platform with a local at Hillsdale and then connecting cross-platform with a Fremont-Livermore-XXX shuttle at the next stop, and vice-versa (time-direction symmetry) northbound.

    ReplyDelete
  33. BuceMcF: "Dedicated HSR tracks as a CHSRA planning position is easily understood as a way to avoid having to think through problems. However, they have to think through those problems in any event with phased implmentation."

    Joey said...
    "And you wonder why some of us don't trust them?"

    Because they exist in the real world? Every organization has a bias against doing additional work without clear reward to the organization. Sometimes the individual instinct of workmanship of the people working on a project leans against this, but it would be silly to be "surprised" when an organization acts that way.

    ReplyDelete
  34. Adirondacker1280002 May, 2011 11:46

    it would be silly to be "surprised" when an organization acts that way.

    Especially an understaffed underfunded one. They should have gone to 37 hour days, 9 day weeks and 502 day years!

    ReplyDelete
  35. It becomes suspicious when some of us, sitting behind monitors in our free time, can come up with better solutions than they can.

    ReplyDelete