Van Ness BRT: Why We Need It

Last week the SFCTA held two scoping meetings to get public input in drafting the EIS/EIR for the Van Ness Bus Rapid Transit (BRT) project. This is Part 1 of a three-post series that combines the findings from the earlier feasibility study, the discussion from the meeting, and some of my own comments on the current conception of this project:

• Part 1 | Van Ness BRT: Why We Need It
• Part 2 | Van Ness BRT: Design Alternatives
• Part 3 | Van Ness BRT: Service Plans

At the meeting, rather than filling out the comment card with comments, I wrote the URL of this blog. Rachel Hiatt and other BRT team members at the SFCTA should be checking up on these posts. Since this is a public scoping period, I’m sure they would appreciate your input as well, so please leave your own comments and ideas about this project in the Comments section!

In some sense, this first post is simply introductory, but I thought it would be helpful to start by laying some facts and figures on the table, all of which confirm what any regular transit rider on Van Ness Avenue already knows – namely, that buses are slow, crowded, unreliable, and bunch up too much. In fact, average bus speeds on Van Ness are between 5 and 7 mph (which is even slower than Muni’s incredibly slow system-wide average speed of 8.1 mph) and buses in this corridor spend a full 50% of their time stopped. When dwell times (associated with picking up and dropping off passengers) are subtracted from total transit time, on average, buses still travel as much as 35% more slowly than cars, which suggests that dwell time alone does not explain why buses are slower than cars. Rather, forcing buses to run in mixed flow with cars has a nontrivial detrimental effect on bus travel times and reliability.

During afternoon peak hours, headways on the 47-Van Ness and 49-Van Ness/Mission are such that a bus from one of those two lines should come every 3.5 minutes, on average. If this actually were the case, it would be great, especially for those people who are traveling only on Van Ness itself, and who could hop on either a 47 or a 49. In reality, though, you could very well end up waiting a lot longer, and when the bus finally does come, there are at least two in a row, because some traffic incident has caused bunching. One figure from the feasibility study suggests that even if roughly 70% of 49 buses started their southbound runs on time, only 41% of the buses were on time by O’Farrell. This distance is only 1.5 miles, a small fraction of the entire 49 route. On average, 4% of 49 buses were bunched by O’Farrell, and 8% by Oak/Market. That figure climbs up to 22% at afternoon peak hours, if we include both the 47 and 49 lines. This reflects the not-uncommon observation of a 47 and 49 bus pulling up to a stop at the same time.

Speed and reliability issues all come down to the fact that buses are forced to run in mixed flow with automobiles. Not only are buses stopping at stops and traffic signals, but they are also stopping every time a car slows down for a turn or pulls out of or into a parking spot. The fundamental difficulty is that we are treating high occupancy vehicles – a bus, with 60-90 passengers – at an equal level of priority as a single occupancy vehicle. Buses are so slow and are given no priority, so drivers have no incentive to leave their cars at home. One of the great potential benefits of BRT is that, in addition to improving wait and travel times, we also have the opportunity to increase transit ridership in this corridor.

Two-thirds of all trips in the Van Ness Corridor are local (that is, they begin and end in San Francisco), while one-third are regional. The vast majority of these regional trips either begins or ends in the city; only about 0.5% of trips via Van Ness both start and end outside of the city. Roughly 25% of trips in the Van Ness Corridor are by transit. Interestingly, this figure is about the same for local and regional transit trips; in fact, the regional transit share is even slightly higher than the local. This means that roughly 40,000 daily trips (which accounts for 50% of the total trips in this corridor) are automobile trips that begin and end in San Francisco. It is especially this group of drivers that should be targeted in an attempt to increase ridership and transit share.

At any rate, the only way to begin solving the problems discussed earlier (short of constructing an expensive subway) is to completely separate bus traffic and to give buses priority at traffic signals, but there are different ways that this abstract idea could translate into actual designs. The designs currently being considered will be the subject of Part 2.

Above image courtesy San Francisco County Transportation Authority.