11 February 2014

Draft Business Plan Impressions

The high-speed rail draft 2014 Business Plan is out, with some interesting implications for the peninsula corridor.

Example service plan, from the 2014
draft business plan supporting documents
New Emphasis on Regional Markets

The ridership model has been (somewhat) updated, although much more is supposed to come, and it's clear that there is now more emphasis on serving local markets.  This theoretically places the high-speed rail system in direct competition with Caltrain.

An example service plan, buried in the supporting documents, is shown at right.  The majority of early-morning trains are local commuter milk runs, including 4 trains per hour serving peninsula stops as well as 4 trains per hour from Palmdale to Los Angeles.

That's all well and good for starting up service in the morning, but one wonders how these customers are catered to during the evening rush (especially to Palmdale) when these commuter milk runs will have to co-mingle with long-distance express trains long before the midnight shut-down, unlike in the early morning when the milk runs coincide with the 6 AM startup.  Here on the peninsula, the speed differential between HSR and Caltrain will be far less, so this problem isn't as acute.

Realistic Peninsula Corridor Trip Times

The trip times embedded in the business plan and the underlying ridership calculations now appear consistent with what can be achieved in a Caltrain - HSR blended service pattern on the peninsula corridor.  These trip times are significantly longer than the 30-minute theoretical SF to SJ minimum that could be achieved under ideal conditions as demanded by the HSR bond measure.  While this shortfall is sure to be the subject of further litigation, it should not detract from the far more egregious loss of trip time due to poor routing choices in the southern part of the state.

The trip between San Francisco Transbay and San Jose is conservatively timetabled at 47 to 49 minutes including a two-minute stop in Millbrae, which is consistent with Caltrain's analysis of blended Caltrain / HSR operations assuming a maximum speed limit of 79 mph for all trains (the same as today).  This isn't too far off from today's Baby Bullet timetable, which includes three more stops than assumed for HSR and therefore takes an additional 9 minutes.

Unrealistically Low Peninsula Fares

The fare model used to estimate ridership and revenue within the SF Bay region is based on an extremely simplistic MTC model that sets the fare to $14.97 plus $0.1283 per mile (in 2013 dollars).  This results in one-way fares of $22 from SF to SJ or $22 from Gilroy to Redwood City.  Such intra-regional trips displace higher yielding long-distance trips, since each seat sold for SF-SJ is a seat that cannot be sold for SF-Los Angeles.  Under any realistic scenario, the HSR operator will manage yields and set a fare structure that discourages low-yield, short-distance trips.  Look no further than the Northeast Corridor for an illustrative example: for a trip from Stamford, CT to New York City (roughly similar in time and distance as SF to SJ), a seat on Amtrak's high-speed Acela Express typically costs well over $100 one way vs. $14.50 on Metro North commuter rail.

Using these ridiculously low fares, the ridership and revenue model for the year 2040 predicts 2.5 million annual HSR riders within the SF to Gilroy corridor, worth $51 million / year in revenue.  That's approximately one-sixth of Caltrain's ridership today, but about two-thirds of the fare revenue.  The model essentially predicts that HSR will poach a large portion of Caltrain's highest-yielding passengers.  Caltrain need not fear for their revenue, however, since the simplistic MTC fare model that underlies this result is not likely to be used by a real-world, for-profit HSR operator.

To Redwood City or Not to Redwood City

Most of the business plan assumes only three peninsula corridor stops for HSR: San Jose, Millbrae and San Francisco Transbay.  The ridership and revenue memorandum, however, includes Redwood City in a fare table.  Furthermore, some of the peninsula trip times in the service planning memorandum are timetabled at 49 minutes (see example service plan, page 8, making sure to account for arrival vs. departure times), which is consistent with an additional stop at Redwood City.  The 33-minute trip time indicated between SJ and Millbrae is very conservative and could easily be timetabled at 29 minutes (departure to arrival) with a 79 mph speed limit and including two minutes of timetable padding.  The extra stop in Redwood City would bring it up to 33 minutes.  Coincidence?