Van Ness bus rapid transit. Courtesy of SFCTA.
In San Francisco, the Transportation Authority and SFMTA are moving forward to recommend a Locally Preferred Alternative (LPA) for bus rapid transit on Van Ness Avenue, signaling that a long planning process will be drawing to a close later this year. This critical north-south corridor between Mission and Lombard is served by Muni’s 47 and 49 routes, as well as Golden Gate Transit, and offers transfers to many east-west Muni services to downtown and the west side of the city. Although the 47 and 49 offer frequent service on paper, actual headways do not reflect these theoretical combined headways. Buses bunch frequently, and the mix of 40-foot and articulated buses operating on the corridor can create a mismatch between the size of the crowd standing at the stop and the capacity of the vehicle that arrives to pick up that crowd.
Although not a cure-all because of the considerable portions of both routes that will operate outside of the Van Ness facility, the package of amenities offered by BRT — particularly dedicated lanes, signal priority, and more widely spaced stops — will help improve these conditions. Toward this end, the TA and Muni are recommending a real BRT alternative, with dedicated transit lanes in the center of Van Ness and stations placed to allow for transfers to intersecting routes. The stations will allow boarding on the right, avoiding the need to procure special vehicles opening on both sides, although lanes will weave slightly around parts of the existing median. The recommended LPA also promises to improve transit performance further by eliminating nearly all left turn opportunities for automobiles in the corridor.
Center-running LPA recommended by staff,with most left turns eliminated. Courtesy of SFCTA.
One reason why it is particularly gratifying that Muni and the TA are endorsing a center-running LPA is that, until the announcement this week, it was not necessarily clear that a center-running design would be recommended over the inferior side-running alternative.
From the perspective of improving transit performance, building dedicated bus-only lanes in the center of Van Ness (as depicted above in the sample schematic for the recommended alternative) is unquestionably the winning formula. While the side-running design would benefit from improvements to stops, it would also allow motorists to enter the marked transit lane for the (officially stated) purposes of parallel parking and right turns, and thus would practically be only marginally better than the so-called “transit only” lanes already in place in certain corridors. On the other hand, buses operating in a dedicated space free of traffic, vehicle turning movements, and parking maneuvers will perform better even than buses relieved of a subset of those conditions. By 2015, bus speeds would increase on average from the current 5 miles per hour to 7 mph, but only 6 mph for side-running. Center-running transit travel times in the corridor would decrease 33 percent with limited left turns, but just 19 percent for side-running. Center-running also performs better on reliability, increasing ridership, and reducing operating costs (or, alternatively, allowing service to be increased at no additional cost). [1]
Van Ness BRT proposed stations at Mission, Market, McAllister, Eddy, Geary, Sutter, Sacramento, Jackson, and Union (south to north). Courtesy of SFCTA.
Originally, an LPA for Van Ness was supposed to be announced a couple months ago, but it was not actually announced until this week. A reason for the delay was a disagreement between Muni and the TA regarding which alternative should be moved forward as the preferred alternative. Although not aired and debated in a public forum, this disagreement centered on misgivings expressed by Muni about adopting a superior center-running alternative. This was not the word of the agency, as the MTA Board has yet to weigh in on the LPA, but rather, misgivings expressed by certain individuals at Muni.
It may seem counterintuitive that a transit operator would be reluctant to endorse an alternative that offers the best opportunity to improve reliability and transit travel time on one of its most important corridors, reduce operating costs, and increase ridership. On the other hand, basically any time I had occasion to discuss Van Ness with TA staffers on an individual basis throughout the study process, they came as close to favoring a centered transitway as one could reasonably expect from individuals working for an agency that has not yet completed its review or taken an official position on the issue. Once it was confirmed that a disagreement requiring resolution explained the delay in selecting the LPA, it was not all that surprising that Muni was at the root of the discord. Less clear is why. Innate distrust of buses with doors that open on the left? Pressure to institute impromptu stops between branded, more widely spaced “rapid” transit stops? Paralysis by fear of the unknown? Simple agency inertia? A fear that transit speeds might actually catch up to those of a century ago?
It is an unfortunate reality that, among the modes that might be considered for a transit upgrade, BRT is particularly susceptible to dilution in the face of political weakness. Even for Van Ness — a critical transit link, where the City is taking advantage of the street overhaul to complete other public works improvements in the corridor, including resurfacing — the urge to dilute and reluctance to commit can come into play. It is therefore encouraging to see the two agencies arrive at mutual agreement to do what’s best for transit riders.
The LPA for Van Ness BRT will be considered by other boards and commissions throughout May, including the Planning Commission on May 10, the MTA on May 15, the TA’s P&P committee also on May 15, and the TA on May 22, with the latter three presumably voting on the recommended alternative. With Muni and the TA uniting to endorse a center-running project, the essential features of this recommended alternative have a good chance of moving through the approval process unscathed, thereafter to be analyzed in the final environmental document and hopefully officially approved later this year. San Francisco is ready to put transit-first into action, by reallocating street real estate for exclusive use by transit riders and committing to build real BRT on this important corridor.
—
[1] Measuring reliability as the probability of encountering an “unexpected” stop (i.e., a stop forced by traffic conditions or traffic signals, rather than to drop off and pick up passengers), the TA estimates that buses have a 30 percent chance of encountering an unexpected stop on each block for center-running BRT with limited left turns, compared to a 50 percent chance for side-running. The TA also estimates a 37 percent increase in ridership on routes using the BRT facility for center-running (compared to 29 percent for side-running).
Transbay Blog 
It’s important to reserve the most important corridors in the city for: c) buses.
Posted by Peter Smith (@shmooth2) | 30 April 2012, 8:39 amThat hardly seems like an improvement. Going from 5mph to 7mph is minor, or are they talking about average speeds, including time spent at stops? If Amdahl’s law is at work and the time at stops is the limiting factor, you would still expect overall speed to increase dramatically due to the reduced number of stops, and with traffic light priority the buses should be able to drive at 25mph. As for still having 30% probability of an unexpected stop, what accounts for that if left turns are axed?
The single most effective way to improve Van Ness corridor performance would be to introduce a new bus line that takes Van Ness from Fort Mason down, then turns onto Market to the Ferry Building.
As for Muni’s resistance, perhaps it has to do with the fact the BRT will only highlight just how many runs are canceled because 12% of drivers take undue “sick leave” without facing any kind of disciplinary proceeding:
http://www.baycitizen.org/transportation/story/next-bus-never/
I tried taking the 47/49 when I lived at Van Ness & Sutter and commuted to 2nd & Harrison. It took 45 minutes each way on average (connecting with the 27 at 9th & Bryant), vs. 15 minute by car. Paying an extra $200-$300 a month to get an hour of my life back was a no-brainer. Similarly, now that I live in Forest Hill, it takes me an hour each way by Muni, despite near ideal conditions (direct service via the T-Third, no connections), vs. 30 minutes by car. Muni is simply not competitive, and it seems MTA has taken the approach that since it can’t improve Muni performance, it will handicap cars instead, to force transit in true stalinist style.
Posted by fazalmajid | 30 April 2012, 8:55 amI’m certain that 5 to 7mph refers to average speeds, including stops. As for what accounts for the unexpected stops, perhaps the signal priority isn’t 100%, in that it can adjust the timing of the lights, but not radically, so that the buses still have to stop if they’re unlucky. That’s just speculation, though.
For Van Ness & Sutter to 2nd & Harrison, google maps lists 24 minutes on the 2 to the 10, though you may have had a reason for not taking that route. It also offers the 10 and the Golden Gate Transit 101, which goes down Market and up Van Ness before heading north. I imagine this is a case where technology could help a lot, by locating the best line at any given time, without requiring the rider to be an expert on the system.
As for “stalinist style”: Muni carries a LOT of passengers. The idea that making dedicated lanes is just a plot to annoy drivers is ridiculous. Indeed, dedicated lanes open the door for much higher passenger numbers on these streets, with good traffic times, while leaving them as is sees them jammed even with today’s traffic.
Posted by Alai | 30 April 2012, 1:26 pmIt looks to me like the biggest downside of this setup will be the constant swaying back and forth, making the ride unpleasant at best and limiting speeds. Wouldn’t it be better to move the platforms–even by just a foot or two, maybe by squeezing the wide car lane, or even by removing a small slice of sidewalk?
Posted by Alai | 30 April 2012, 2:09 pmWhy “distrust” left door buses? you ask. Because neither the 47 or 49 run their entire routes on Van Ness and NO GGT buses have left hand doors. Given Muni’s ‘stellar’ maintenance record, needing a specialized fleet for Van Ness is dumb.
Posted by david vartanoff | 1 May 2012, 2:01 pmDavid, I hoped it was clear that I was being tongue in cheek by that point in the piece, but maybe it wasn’t that clear after all. Clearly there are several disadvantages to having a special portion of the fleet just for Van Ness, beyond what you’ve mentioned (although, if they did decide to pursue that alternative, one would hope eventually it would be used on corridors other than Van Ness). Also, the vehicles wouldn’t just be “left door buses” — you would need doors on both sides to cover the portions of the routes outside of the BRT facility.
In any case, given that center-running can be implemented without a specialized fleet, the specialized fleet is not a reason to oppose center-running on principle.
Posted by Eric | 1 May 2012, 2:10 pmActually I missed cheek. Having just been to the AC rubber stamp meeting for their terribly designed BRT which does plan special two side door buses, I may not have been alert. Van Ness is a good place for dedicated lanes in my view; it is slow as molasses most of the time.
Posted by david vartanoff | 1 May 2012, 7:48 pmI do wish GGT would be allowed to pick up and drop off passengers within SF. This facility would make a strong case for that kind of cooperation, but from my understanding it would take a change in state law to do it.
Posted by David Edmondson | 1 May 2012, 9:33 pmEric, glad you got this important discussion going with your great analysis. Also, while it’s incredibly disappointing that BRT in the East Bay won’t go to Berkeley (Berkeley’s fault), the project otherwise has a very strong design. David V., please expand on that.
Posted by Marta | 2 May 2012, 11:28 amThank you for writing about this, Eric. It looks like a very good design. It’s frustrating to see how long it takes to just design and approve such a project, let alone build it. It’s admirable how some activists have the patience and staying power to stay with such an undertaking for so long.
Posted by Sprague | 2 May 2012, 1:03 pmReally interesting, glad to see something moving forward! I rarely take those lines, but they have sometimes been terribly overcrowded, at the evening hours when I do use them. “Why not eliminate more left turns?” is a good question–after all, three rights make a left, and 19th Avenue has a good long stretch without any left turns.
Posted by Al_Pal | 20 May 2012, 4:20 pmjust add this until later when light rail can be added to underground services as well as geary an making the n_juda underground to 19th ave
Posted by tyrone wright | 23 May 2012, 8:25 amJust chiming in that i’m also happy to see real BRT implementation. should give a shot of calcium for those weak-spined berkeley merchants and parker porkers.
Can’t wait for this to also happen For Reals on geary street!!
Posted by newoaktown | 17 June 2012, 9:26 pm