What’s the Best Design for Van Ness BRT?

The best choice for transit riders comes down to two center-running options for Bus Rapid Transit on Van Ness Avenue. Images courtesy of SFCTA

After years of delay, the 2016 target date for the Van Ness Bus Rapid Transit project seems more tangible than ever. The San Francisco County Transportation Authority recently released its draft environmental impact report and will select one of several proposed design alternatives in the spring.

The SFCTA is asking for public input on the different options and the draft report, which includes a trove of information for planners and transit advocates to consider when weighing each design.

Last week, the San Francisco Transit Riders Union’s Rapid Transit Working Group met to discuss the alternatives.

“Ultimately, we’re looking at what is going to create the best, 21st-century riding experience for transit riders on Van Ness Avenue,” said SFTRU board member Rob Boden. SFTRU members are considering which design to endorse, but the organization hasn’t taken a stance yet.

The group’s top priorities, said Boden, are improving transit reliability and passenger comfort. The EIR analyzes those factors along with everything from median widths and greenery to bus weaving.

Bus Rapid Transit has appeared in a variety of forms in cities around the world, but generally includes amenities like dedicated lanes, pre-paid ticketing, all-door level boarding, and limited stops that feel more akin to riding a rail system than a bus.

For the Van Ness project, the most significant question boils down to this: Should buses run between a pair of medians or alongside a single center median?

A comparison of bus and auto speeds for each of the project's design alternatives.

The EIR analyzed three design alternatives (plus a status quo scenario), but only the two options that place bus lanes in the center of the road appear to maximize the project goals of improving bus speeds, reducing operating costs, and increasing ridership. Each of those options carries some pros and cons.

On one hand is the dual median alternative, which would separate a two-lane busway from other traffic by placing it between two medians, where passengers would disembark from the right side of the bus. One problem with this design, the report says, is a risk of delays and head-on crashes when one bus needs to pass one another. The price tag for construction ($130 million, compared with $119 million for the other center-running alternative) and maintenance of the busway and landscaped medians is also higher than the others, and it would require the highest number of median trees removed.

Alternative 3: a center-running busway between two medians.

The other center-running alternative would place each bus lane on either side of a single center median. Since boarding at BRT stops would be limited to the left side, buses would require doors on both sides to use the platforms (buses would still need right-side doors for the remainder of their routes). This could limit Muni’s flexibility with its vehicle fleet, and buses operated by Golden Gate Transit would only be compatible with one stop on the two-mile stretch (between Geary and O’Farrell).

Alternative 4: center running bus lanes on the outside of a single median.

Tilly Chang, the SFCTA’s deputy director for planning, said staff has consulted other municipalities that have used BRT vehicles with doors on both sides, including in Cleveland, Ohio and Eugene, Oregon.

“We found that they have not experienced significant incremental operations and maintenance costs versus their right-door only buses,” said Chang. “However, this requires further study before assuming a similar experience would be the case here, and we are undergoing that exercise right now in as much as it can inform the decision-making process.”

Another concern raised by SFTRU members about the design is the lack of a physical barrier preventing other vehicles from entering the bus lanes, but Chang said the lanes could be slightly raised.

On the upside for the alternative are wide refuge medians for pedestrians, added separation between the boarding platform and auto traffic, and the quickest and least intrusive construction period of any of the alternatives.

For both center-running alternatives, it’s also worth noting that planners have the option to prohibit all left turns on the corridor (save one in each direction at Broadway and Lombard). This measure, called “Option B”, would maximize the project’s benefits in virtually every criteria, including lowering operating costs from $6.1 million to $5.6 million and reducing vehicle crashes, according to the report.

Alternative 2: side-running bus lanes.

The third design alternative, which would place bus lanes on alongside parking lanes, would yield the poorest results for transit performance, according to the report. Bus speeds are projected to increase only about half as much as the center-running designs, and this configuration would carry the highest operating costs.

That’s no surprise, given the lanes would likely see frequent incursion by other vehicles. The problem can already be seen on existing bus lanes on Mission, Geary, and O’Farrell Streets, where drivers frequently stop in them to make right turns, pull into parking spaces, and double park. As a result, passengers are subject to constant delays as buses weave around other vehicles.

Colored pavement treatments, included in each of the design alternatives, could mitigate that effect by discouraging drivers from entering the lanes, but it’s no substitute for center-running, separated bus lanes.

The SFCTA’s next presentation on the project will be an online webinar on Monday, December 5th at noon.

To find other meetings, learn more, and submit comment, see the SFCTA website. You can also get complete details on the design alternatives in the EIR’s Alternatives Analysis chapter [PDF].

  • The top option is reminiscent of the Santa Clara Caltrain station which is being redesigned at the cost of 100’s of millions of dollars. If both buses arrive at the same time, people boarding the SB bus must wait for the NB bus to clear before they can board. Bad.

  • I do hope they choose something that would be compatible with GGT as well.  No fewer than 6 routes use Van Ness: 10,70, 80 and 101 all day, and 93 & 97 for commute times.

  • I could be wrong, but I thought that the oncoming bus in that top photo has its own  right-side platform on the opposite side of the intersection, out of our view.

  • Murph, I guess I’m not following. From the picture above I don’t see why the SB bus and NB bus would ever have to wait for each other if they both have their own boarding platforms (as throgers notes below)?  Seems like lots and lots of rail stations all over the world operate this way? I would say given that this design would meet GGT’s needs best as well and would keep cars completely out of the bus lanes, it looks to be the best solution.

  • Jarkatmu

    The platforms are staggered in that top image. NB platforms are on one side of the intersection and SB platforms are on the other side of the intersection. This VTA light rail station reflects the essence of the staggered platform proposal.  


  • Jason

    Take Alternative 4, run the buses contra-flow, and add a small lane barrier (a curb a few inches wide).  Now right-side doors work and people won’t dare drive in the bus lane (they’d be going the wrong direction, and have to drive over a curb to get to it).

    Contra-flow dedicated bus lanes seem to work fine up in Oregon, and definitely get the message across that THIS LANE IS NOT FOR CARS.

  • Andy Chow

    I am not a fan of the center platform. That requires two separate fleets (trolley and hybrid buses with left side doors) which would have operating and reliability impact. The other downside is that it would reduce seating capacity.

    Sharing the busway with Golden Gate Transit is a good thing. However Golden Gate Transit buses tend to have really long dwell time because riders board the bus one by one and feed the dollar bills in one by one. To make GGT buses not to delay Muni service on the busway, they should sell GGT bus tickets through the ticket vending machines at the station. That way they could cut the travel time further (when their travel times are quite long already).

  • David

    Standard right side door buses whether dual platform or counterflow.   One for fleet flexibility, two for GGT, three because custom fleets cost extra to buy but don’t last any longer or perform any better   
    As to GGT, they are already moving to Clipper and will likely be cashless by the time anything is actially built.  

  • Tim

    Its really not worth doing (and then wouldn’t technically be a BRT system, just another bus lane disaster like Mission St.) unless you have a physically separated bus lane. So option 3 it is. I like Andy Chow’s idea to put Golden Gate Ticket Machines (or incorporate them into the Muni ones) to speed up boarding.

  • Keithsaggers

    re. GGT on VNA
    EIR Chapter 10,
     80% of existing GGT passengers use either the stops atGeary/O,Farrell or a stop that is not located on VNA
    EIR Chapter 2,
    Alt. 4 GGT could have an additional stop @Chestnut or Union Sts..

  • Andy Chow

    I think there could be a combination of alt 3 and 4. Basically retain outside boarding platforms where left side door vehicles will not be necessary, and retain center median (rather than two medians) for blocks that don’t have stations.

  • Anonymous

    I think the side-to-side motion that would be required would discomfort passengers and limit the speed of the buses.

  • Chris P

    Staff working on this project have told me that option 4 would include a curb physically separating the bus-only lanes from the mixed-traffic lanes.  Unfortunately, the visual simulation doesn’t clearly show that curb.  If option 4 didn’t include substantial physical barriers, I agree it wouldn’t be worth doing.

  • kg

    Alternative 3 appears to require ripping out the beautiful old eucalyptus trees, and planting palm trees that HAVE ABSOLUTELY NO PLACE IN SAN FRANCISCO. As you can tell, I’ll lose my mind if they put in palms. This ain’t SoCal people! Not to mention, it’s a contradiction of the city’s general plan for Van Ness Ave.

    In that sense, Alternative 4 seems to be the most reasonable with the least amount of environmental impact.

  • TwinPeaks_SF

    I absolutely agree, but palm trees are being considered because they are one of the only trees that will clear the overhead wiring for the trolley buses. That’s why they’re on the Embarcadero and Upper Market. Streetsblog readers – do we know of any alternatives?

  • Anonymous

    You’re ridiculous. Palm trees have as much of a place in SF as Eucalyptus which is also a non native species. Embarcadero is beautiful – and the palms require much less maintenance then a normal tree. Look at the pictures of SF in the library archives – they are all online. You’ll see palm trees even in the oldest pics of SF – before most of the places in Socal that have them now were even cities.

  • Matt Baume

    Eucalyptus aren’t native either.

  • Anonymous

    @77345bdee4b30563c178c37b8f3d377c:disqus  I totally agree that they shouldn’t be putting in palm trees: they are not native to the Peninsula and always look brown and wilted. And I agree that it’s a relic from several generations ago when they wanted to try and pretend like SF was some tropical location. We are not the tropics and we should be embracing our own native climate and species and stop trying to imitate something else.

    But I disagree that we should plant eucalyptus trees since they are also not native.

    I think this whole issue of planting native trees is actually quite underrated, since not only do native species look better since they by definition thrive in this climate, but they require no irrigation, little maintenance, and attract native birds, insects, and animals. And I believe the last issue is highly underrated, as native trees and hence native birds, insects, and animals give a city a vitality that innately appeals to humans and hence makes a city much, much more livable than most people realize.

    @twinpeaks_sf:disqus There are plenty of other native tree species that can be used instead of palm trees 9 (note that this list includes not just trees but shrubs and flowers):

    not all native trees are appropriate for the urban environment, but
    that’s why the city should be working with people like Nature in the
    City (http://natureinthecity.org/) to
    make sure appropriate native plants are used throughout the city. I’m positive there are trees that can be used on Van Ness that are native and work well there. The
    point is: there are many options other than non-native palm trees,
    eucalyptus trees, etc.

  • Anonymous

    Wake me when the discussion turns to rail on Geary.  Oh wait, what am I saying?  We all know that’ll never happen.

  • Anonymous

    The canary island palms are all brown looking and wilted in SF? huh.

  • kg

    To all who read this as me advocating planting of new eucalyptus, that’s not what I’m saying. That is what the city’s plan calls for. I’m merely stating that I’d hate to see what is already there (since 1870 if what I read is accurate), be destroyed. I do believe it’s the city’s job to identify something that’s both native and functional.

  • BBnet3000

    Alternate 3 with the platforms for each direction after the intersection seems like the best option for using the narrow ROW, having truly seperated center lanes, and keeping right-side doors.

  • Callingcarole

    Palm trees yes, everything else NO…where will the existing gridlocked traffic go?

  • Sean

    I like the single center median, but having a dedicated fleet is very problematic.

    I dont know why everyone says that the Geary GGT stop is the most important, that is just false. I see more people at Union/Sutter. With the 101 express route, I think that GGT could consolidate stops so there is only a stop at Sutter/VN, and Broadway/VN. No lefts except Broadway and the Bush/Pine couplet. There could be a traffic engineer/driver revolt otherwise. The 70 route can take the old lane.

  • Kevin

    I agree, just make the lanes contra flow, and you can run existing buses on this BRT line.

  • TwinPeaks_SF

    As is the whole point of BRT – hopefully much of the intra-SF traffic will move to Van Ness BRT as opposed to driving.

  • Anonymous

    Ah yes, the question of the ages. Where did all the traffic go when they got rid of a lane on Valencia? Where did all the traffic go when they banned most of it from Market St? Where did all the traffic go when they tore down the Embarcadero Freeway after the 1989 earthquake?

    Though it usually gets worse temporarily immediately after road diets, drivers adapt and in the long-term they either are more efficient with where they drive (ie, they drive less) which usually includes taking public transit, walking, or cycling or else changing where they live or work so they don’t have to drive so much. The reality is, the people will adapt to whatever you put there, so it makes sense to put the most efficient, most healthy, and least environmentally destructive modes of transit in place and watch as people start using them.

  • Aaron Bialick

    The EIR’s Transportation Analysis [PDF] does a pretty good job of addressing this.

    Generally, car congestion is projected to be worse without BRT than with it.

    Here’s an excerpt:

    “Van Ness BRT is the primary transit street in the corridor, as opposed to Franklin and Gough streets, which are the primary private vehicle streets. BRT would help Van Ness Avenue function more efficiently and increase transit ridership.

    Vehicle diversions to all other streets in the corridor would add up to less than 7 vehicles per minute under the build scenarios.”

  • Anonymous

    Contra flow would mean that buses traveling in opposite directions would have to cross paths when entering or exiting the BRT corridor, at Lombard St and Mission St. Sounds like a nightmare to me. In any case alternative 4 will apparently have curbs, not shown on the image, to prevent cars from entering the bus lane.

    Finally, a new fleet of buses is going to be purchased by Muni anyway to enable level boarding- I’m guessing they will be low-floor (around 15 inches) with platform height to match. There could be one extra step that slides out from below the bus at regular stops without a platform so that the step isn’t so tall. These buses won’t be able to climb steep hills so will be restricted to the BRT routes and other flat routes. Level boarding is well worth having (think about how much time you spend on the bus waiting for disabled and elderly passengers to board) so the cost of buying new buses and the inconvenience of a bit of operational inflexibility is well worth it, IMHO.

    If GGT don’t want to purchase new buses they may be able to use their current ones with alternative 3, they just won’t have the advantage of level boarding. With alternative 4 they will definitely need new buses with left-side doors.

  • Mario

    Option 4 is supposed to have a raised lane like the N. Of course cars can still use it, but it will be pretty obviously the wrong thing to do.

  • Ty

    For the BRT line’s sake, I hope they don’t. GGT busses take forever to load and unload along Van Ness because almost everyone pays cash fares, which will delay BRT busses significantly. If GGT patrons will be subject to the same POP service and are able to buy their tickets on the platform before entering the bus, then it would be fine… but if that’s not the case, I’d rather have the GGT busses relegated to the auto lanes to keep the BRT lanes for true BRT.

  • Grrlfriday

    The best design for Van Ness “BRT”? That’s easy. It’s the one option no one considered: surface light-rail right out of the gate.

    The likelihood is that once all the money is spent to build “rail-ready” infrastructure, we will come up with excuses for decades to come as to why we shouldn’t ever get around to converting all this expensive new infrastructure to actual rail. Once you factor in how expensive the “BRT” infrastructure is to build, and how much more buses cost to operate than trains over time, buses are far more of a money-sink than trains, even the so-called “rapid” ones. And no one likes riding them as much, so they will never attract the ridership. Just look at the overwhelming popularity of the F-line down Market Street, and ask yourself if it would have nearly as many riders if it was only buses.

    The tracks through the tunnel to Fort Mason at the northern end of Van Ness, and the F-line tracks through Fisherman’s Wharf could both be connected to a rail line up Van Ness which divided into two spurs at its northern end. That would be a rail line worth riding. A “rapid-bus” system up Van Ness would certainly be better that the current alternatives, but it will never justify the expense, and in the end we will still be left with a system no one really loves. I would love to see an alternative here that is not just less mediocre, but is great public transit that everyone wants to use. To me, that does not say “buses”, it says “trains”. It’s so frustrating that that option was never seriously considered.

  • center running seems to be the obvious choice, but i don’t think left turns should be eliminated the whole route, i’d imagine that would just increase the number of cars on the smaller streets as people make three rights instead of a left.

  • i totally agree! there is the ridership potential and that part of the city isn’t served by the muni metro

  • sohcrates

    I am very sympatheitc to the idea of incorporating native species into our public works projects. However, from a standpoint of public safety, and at least for a major thoroughfare like VanNess, palm trees make sense because they grow vertically, and would obscure the roadway less than a native species that has a lot of horizontal growth, like a manzanita or a native oak. 

    Redwoods are a native that have the vertical form going for them, yet as much as I may also love them personally, they don’t really work well in urban environments because they are very aggressive growers, especially with their wide reaching roots. A couple years ago we had to call out a plumber to our house in the sunset to replace our sewage drain because the roots from the redwood that was planted in our neighbors’ yard had broken into it and wouldn’t let anything pass through….. not a very pretty site. I can only imagine how much uglier something like that could get in a denser place, like the Van Ness corridor.

    If you want to talk about places where the native touch could really work: Civic Center and UN Plazas could really be livened up that way. Justin Herman Plaza would look great with some native grasses and shrubs, and now is the time to do it, since that grassy median (that nobody ever used anyways) has been trampled up by the occupiers. The grassy strips on the terraces at Union Square would look much more inviting if they put in a treatment similar to the one that was put in on the Guerrero Median.

  • Out in DC, the plan for the K Street Transitway has center-running bus/rail lanes, giving fleets the flexibility to move in and out of the dedicated lanes where appropriate while also providing the increased capacity of a light rail line.

    I’d much rather see the bus lane extend out on Lombard to the Bridge, and get something going for Geary.  Actually, as a Marinite I’d rather see bus lanes going across the bridge and up as far as the Spencer Ave. Bus Pad (we’ve got 21 lines moving at rush hour!), but that’s not gonna happen.

  • Sean

    Maybe just restricted lefts at peak times of the day like Lombard. But I dont think that Funston and 14th have that much congestion if you look towards Park Presidio as an example.

    As far as TPS, I think the T-Third is a good example of how that isn’t always the answer. I’ve seen a single truck taking a NB left at Folsom block a 2car train packed full of Giants fans. LOS just ruins TPS.

  • Anonymous

    Unlike Geary, Van Ness BRT isn’t even supposed to be “rail-ready”. But then Geary is a better candidate for light rail anyway, as it’s a corridor with 1 heavily used Muni line without any branching (plus 1 GGT line.) Van Ness hosts 2 moderately used Muni lines plus 5 GGT lines, all of which branch at the north and south ends to a bunch of different places. For this reason BRT is better for Van Ness as you can focus improvements on the trunk corridor that they all share. If you upgraded Van Ness to light rail you would need to lay track all the way down Mission to Balboa Park and through Soma to Caltrain, or else have the 47 and 49 terminate at Mission & Van Ness and make people transfer to the new light rail line. Neither of those are desirable solutions.

    For the same reason, BRT is a good interim measure for Geary as you can upgrade the section west of Van Ness to surface rapid transit and keep the section east of Van Ness as is until you find the money for a subway under downtown. East of Van Ness the streets are too narrow and congested for surface rapid transit (bus or light rail) to be effective. Once you’ve built the subway you can put rails on the surface section.

    Also, I reckon the reason they’re planning to end the dedicated BRT lanes at Lombard rather than continuing up to North Point is because they have an eye on extending the T-Third to the Marina using Columbus, North Point (or Bay), Van Ness and Lombard, and they don’t want the Van Ness BRT project to preclude a T-Third Phase 3 project.

  • hopefully the buses can be placed on the outside lanes to make sure SF will never allow bikes on Van Ness. but even placing huge buses in the middle lanes should take care of this. #awesome

  • tstieber

    Canary Island palms aren’t tropical palm species; they are Mediterranean palm species.  Northern California has, by definition, a Mediterranean climate (even though SF has cool summers by the water).  Plants from all Mediterranean climates (Chile, South Africa, Australia, California, and the Mediterranean Basin) all thrive in each others’ regions, so they are appropriate.  If they’re not your taste, that’s legit, but of course they have a place here.  They’ve been here for more than a century.  And all the palms in SoCal are non-native too.  SoCal doesn’t have a tropical climate either, just a somewhat warmer Mediterranean climate than San Francisco, with winter rain and chillly nights.

    Still, I like the existing Eucalyptus on Van Ness.  There are already so many palms in the City these days that it would be nice to keep the leafy character of those trees.  But if the medians don’t allow for much root growth, then palms are a low maintenance choice.  I don’t know why people say they always look brown and wilted.  Some do, but that’s because of fusarium wilt, a fungus that plagues this species all over the state.  It’s not because of climate.

  • Max Canepa

    Alternative 3 seems to be the best viable alternative for all parties invloved. Van Ness adheres to a wide range of mixed uses zoning and also has a significant amount of government and corporate builidng along the streets. What this brings is a mass amount of traffic congestion and promotes people to drive around hopefully to be the lucky ONE to get the spot. Alternative 3 provides a sperated running lane for BRT as well as provides lanes for autombiles and biking.

  • teflonrobg

    The buses in the middle of the road will work better because it works well in the REST OF THE WORLD!


After Delay, SFCTA Board Approves Van Ness BRT Design

A preferred design for Van Ness Bus Rapid Transit was approved unanimously today by the Board of Supervisors, acting as the SF County Transportation Authority Board. Supervisor Mark Farrell, who delayed approval of the proposal a month ago after complaining that he “hadn’t been briefed” on it, said he now stands behind the project after SFCTA staff brought him up […]

SFTRU Forum: Van Ness BRT

From SFTRU: It’s finally happening! “Bus Rapid Transit” is coming to San Francisco! After many years in the making, the SF County Transportation Authority and SF Municipal Transportation Agency staff are recommending an innovative project for center-running buses on Van Ness Avenue that will speed up travel times on the 47 and 49 by as […]

Options for Geary BRT Come Into Focus

Just after San Francisco approved a preferred design for its first Bus Rapid Transit route on Van Ness Avenue, the SF County Transportation Authority showcased the latest conceptual proposals for a companion BRT project on Geary Boulevard. Geary BRT, which has been fraught with delays over the years, is expected to bring relief to riders on Muni’s […]

Supervisor Farrell Delays SFCTA Approval of Van Ness BRT Design

A crucial step in advancing the Van Ness Bus Rapid Transit project was delayed for a month today after Supervisor Mark Farrell, a member of the SF County Transportation Authority (SFCTA) Plans and Programs Committee, complained that he wasn’t comfortable voting on the latest design proposal which he said he “hasn’t been briefed on.” The […]