16 April 2010

Alternatives Analysis Analysis, Part 1

This post is the first in an open-ended series, picking through the recently-released preliminary alternatives analysis. The AA is a voluminous document with plenty of material to be dissected over the next several months, in no particular order.

The Station Access Fallacy

Section 4.1.2 of the Alternatives Analysis discusses track and station arrangement options. In particular, it is pointed out that a key consideration in providing good station access for Caltrain passengers is to consider that the majority of Caltrain stations, and the majority of Caltrain ridership, is on the west side of the tracks. That sounds pretty good in theory... but in practice, consider the following:
  • The rail corridor must be 100% grade-separated, with no crossings, so passengers may only cross the tracks via pedestrian underpasses or overpasses, never at grade.
  • To access an island platform, or the platform on the "far" side of the tracks for at least one direction of a passenger's daily commute, every passenger will use underpasses or overpasses at least twice daily regardless of station configuration, and regardless of which side of the tracks they might live or work.
  • Putting Caltrain on the west side of the corridor, rather than in the middle, might save each passenger a grand total of 30 feet of walking distance, far shorter than the stairs or ramps to the platform or the walk to the doors of an 85-foot rail car.
  • It takes seven seconds to walk an extra 30 feet. Some hike!
In a 100% grade-separated corridor, it is given that station access will be limited to a handful of access points somewhere along the 800-foot length of a Caltrain platform. As long as those are well-planned (e.g. Belmont), that's perfectly acceptable. The station access "advantage" of putting Caltrain on the west side of the corridor is insignificant.

The One Percent Rule

In the AA profile plans of Appendix B, page 3, a design criterion is established: the maximum grade for Caltrain (shared use) tracks with diesel-powered freight trains is 1.0 %. The community impacts of this gentle grade, required exclusively for freight trains, were previously discussed. And yet, Caltrain's own engineering standards state that "maximum design gradient (...) for grades up to two (2)% may be implemented with the approval of the Caltrain Deputy Director of Engineering. (...) Grades exceeding one (1)% shall be limited to tangent length less than 1200 feet." Why is a more stringent standard now being applied? Certainly not for high-speed rail or Caltrain, both of which will easily handle 3% grades.

The results of this conservatism are already apparent. For example, the proposed elevated alignment through downtown San Mateo would require lowering 1st, 2nd, 3rd, 4th, 5th and 9th avenues, with extreme community impact, because (a) the grade is 0.8%, even less than the specification, and (b) there is a requirement to return to grade level before the curve into Hayward Park to avoid having a vertical curve in a horizontal spiral, a track geometry complication exceptionally allowed for (see TM 2.1.2 paragraph 6.1.7) in the CHSRA's own track design guidelines. Let's face it: the age of picks and shovels is over; in the 21st century, computers, lasers and differential GPS should make the construction and maintenance of overlapping vertical and horizontal curves a breeze, without having to demolish half of downtown San Mateo.

With things already tight as they are, the time for making judicious and appropriate design exceptions is now. Using overly-generous specifications will unnecessarily constrict the design space, inflame community opposition, and drive costs through the roof.


  1. But what are the chances of kicking freight off the corridor?

  2. But what are the chances of kicking freight off the corridor?

    Roughly zero. But that's not the point: it's perfectly reasonable to perform a trade-off where, say, a slightly steeper grade with less community impact would be weighed against a train tonnage limit. Right now they are designing by rote to standards which apply just as well to mile-long coal trains in Wyoming.

    Again, the point isn't to kick freight off the corridor, but to make freight meet in the middle with other corridor users and needs. There's a reason this blog isn't called the "Caltrain-freight compatibility blog".

  3. This comment has been removed by a blog administrator.

  4. Yes, Richard, we know you don't like them. Do you have to tell us EVERY SINGLE DAY?

  5. So, Caltrain's existing design standards basically allow for a 2% grade up and down over a grade separation, which seems reasonable enough. Why isn't Caltrain leading the design here, since it seems that they would actually do a better job. Despite all the problems of their existing designs, they're still better than what the HSRA seems to come up with.

  6. Step 1: Decide on a course of action.

    The plan should, naturally, be the same thing that you, your father, and your grandfather have done for the last 100 years.

    A record of continual, decade-in, decade-out failure is a primary attribute of the course of action, as is extraordinary cost.

    "Decision" is too strong a term for what occurs at this step: "just doing the same old crap over and over and over without thinking" is really what happens. You don't get a job as a planner or engineer at Caltrain because you're the best and brightest in the world, after all. Somebody's got to fill the B Ark, and everybody from Japan and Central Europe is too qualified.

    Step 2: Invent "engineering criteria" out of thin air and "vendor prequalifications" that exclude consideration of anything other than that fore-ordained in Step 1.

    Step 3: Context Sensitive Solutions and Public Input. "Thank you for your contribution. Comment acknowledged."

    Step 4: Place the plan from Step 1 out to bid, tailored exactly to those vendors decided upon in Step 1.

    Step 5: PROFIT!

    Step 6: Well, that didn't work out so well. Who could have predicted? Only a quarter the ridership and twice the price we predicted, but exactly in line with "thank you for you input" feedback from people who don't have our long and distinguished record? Who could have predicted that? It must have been due to dot-bomb/global recession/global warning/WMD/something. Anyway, let bygones be bygones.

    Step 7: Repeat.

  7. Adirondacker1280017 April, 2010 13:20

    Step 7 is "Rinse" Step 8 is Repeat.

  8. there is a requirement to return to grade level before the curve into Hayward Park to avoid having a vertical curve in a horizontal spiral, a track geometry complication exceptionally allowed for (see TM 2.1.2 paragraph 6.1.7) in the CHSRA's own track design guidelines.

    There's a good reason for this rule and that is derailment safety. In superelevation transition, one wheel of any bogie is lifted and eventually can lose contact with rail and crest vertical curve makes it worse because it decreases axle load without decreasing side forces, so the derailment safety gets lower. With flat enough transition slopes that must be used on high speed tracks, this problem disappears.

  9. Where is Tomolach with his stupid rants?

  10. Here is a video clip

    from the April 19th, Assembly transportation committee with testimony debunking the myth that HSR systems around the world are profitable.

  11. Testimony ... got any actual reports to back up that guy's claims ... anyone else have any reports to the to the contrary?

    (and I should add that the only HSR system he really mentions is the TGV - the rest of the time he's just talking about transit in general. And SNCF does operate at a loss because of its non-high-speed regional services)

  12. Excluding the locally-funded commuter trains, SNCF is highly profitable, due to the success of the TGV. However, increasing track fees are threatening those profits, so that 20% of TGV services lose money. Needless to say, if only 20% of Amtrak routes lost money, Amtrak would be in the black today. (The local routes are kept separate from the national company, as they should. Amtrak doesn't operate the LIRR, and the Interstate system doesn't operate local streets.)

  13. While Amtrak as a whole loses money per passenger the Acela express (closest thing to HSR in US) makes a profit of $41 per passenger according to Pew Research:


  14. @ Morris Brown

    You can't "debunk" anything when you're peddling bunk.