Sunday, March 8, 2009

How to plan rail service

Those who have been following the California High Speed Rail process will have noticed that there's been quite a lot of recent controversy about the proposed route between San Francisco and San Jose. Leaving aside the propriety of a government authority board member deciding on train service to a station named after himself, there's plenty of worry that the High Speed Rail will bulldoze through towns, destroying homes and leaving huge ugly structures with noisy trains in its wake. Emotions are high on both sides, there's a lot of uncertainty, and not much fact to go on, so arguments are based on differences of assumption, and tend to devolve to name calling. Unfortunately, the High Speed Rail Authority hasn't done nearly enough in the way of planning, mostly because they haven't had money to until now. In particular, I haven't seen pretty much anything in the way of joint planning with Caltrain regarding track sharing, service levels, station design, and improvements to the line. This strikes me as a crucial omission, and I suggest Caltrain and HSRA work together to remedy this at once.

What would a planning process look like? I suggest the following:


  1. Start with ridership forecasts for both Caltrain and HSR. These already exist, the only potential caveat is to not double-count riders when combining them.

  2. Determine a service level that will adequately serve this demand, and make a conceptual service pattern for the line. This has been done on some level for HSR and Caltrain individually, but not jointly

  3. Find the minimal set of improvements to transform the service plan into an actual timetable. This is basically a matter of straightening curves to get sufficient line speed, and providing enough passing tracks to allow express trains to pass locals.

  4. Finally, this process might need to be iterated several times, to account for the fact that ridership is somewhat flexible and it might be much easier to produce a workable timetable if some riders are convinced to move to adjacent stations.


This doesn't seem like it would be that hard. The respective agencies already have ridership forecasts, they just need to combine them and produce a joint service plan, and start devising potential timetables. Caltrain should probably take the lead here, since they own the ROW and will still be the majority operator. And this is something that absolutely needs to get done, to make sure that all the agencies are working with a common set of assumptions. It would be bad if, for example, HSRA was planning for almost exclusive use of two tracks with only occasional incursions by Caltrain Baby Bullets, while Caltrain was planning more significant track sharing, or even a line that wasn't four tracks for the entire length.