BRT Comes Out Ahead of Light Rail, Again

Las_Vegas_BRT.jpgBRT bus in Las Vegas looks a lot like light rail

The debate among policy makers and community stakeholders over the merits of Bus Rapid Transit (BRT) versus Light Rail Transit (LRT) is often heated, and usually centers around whether LRT recoups the substantial capital costs of implementation over time versus BRT, and whether BRT has a more substantial carbon impact. Sometimes it can also boil down to a debate over whether buses are sexy enough to get people out of cars and onto transit.

The World Resources Institute (WRI) recently presented a report comparing BRT
and LRT in the “medium investment” range for the Maryland Transit Administration (MTA) on the Purple Line, which would connect suburbs around Washington DC with the city center.   WRI’s analysis
confirms that BRT is the option that would work locally to fight
global warming, with a medium-investment system cutting carbon dioxide
emissions by almost 9,000 metric tons per year, equivalent to taking
about 1,600 cars off the road (PDF).

In an interview with Worldchanging, the report’s authors, Dario Hidalgo and Greg Fuhs, address the CO2 numbers: "While this could change in the future with a major and permanent shift
to low-carbon energy sources, for the foreseeable future we would
likely continue to see higher CO2 emissions from light rail in this
case," said Fuhs.

The report finds that BRT would
be more cost-effective and lower-risk, particularly with the current economic situation.  With tight state and federal transit budgets, and the recent hullabaloo over the stimulus package, it will be difficult to get substantially more funding for transit.

In San Francisco, Central Subway aside, policy makers appear unwilling to build more fixed-rail systems in the near term, preferring instead to go with BRT.  The SFCTA and MTA are moving forward with the Van Ness and Geary BRT lines, which still need to clear environmental analysis, but could be online by the MTA’s Centennial in 2012.


Because California generates a significant percentage of electricity from renewables and nuclear, the carbon impact would be less severe.  Transit advocates, however, are skeptical of the city’s ability to build and operate LRT to be competitive with BRT.  While the retro trolleys might look good in photos along the
Embarcadero, they are not efficient rail transport compared to Muni LRT
or other systems, and the cable cars are pure tourist sport.

"Other cities are building light rail very cheaply and operating it well," said Livable City’s Tom Radulovich.  "The big problem with San Francisco is that we build incredibly expensive light rail, like the Townsend 3rd, and we run rail like a bus.  We’re not being smart about our investments."

All right Choo Choo Heads and BRT-ophiles, we know you’re reading, tell us what the future of San Francisco’s surface transit should look like.

Flickr photo: NeiTech

  • Does the WRI report include the effects of TOD around stations? It’s not enough to just compare the number of potential riders for a LRT line to a BRT line. If the LRT line induces TOD that’s a lot of walking and cycling trips as well that need to be accounted for.

  • BRT might be best to replace more expensive BART style commuter rail in traditional spoke and hub Cities. That is a BART 4.0 discussion not a MUNI 3.0. But once you get into city centers you need the kind of total grade separation that BRT cannot provide, subways are the only way to go.

    Given the political forces at work, does anyone think that is it remotely possible to find the will to do what it takes to engineer Market Street such that BRT competes with BART between Montgomery and Civic Center? Would that even be technically possible to do, isn’t it mathematically impossible to time signals on a two way street so that transit vehicles never have to stop or to run at speeds that would compensate for those delays?

    My understanding of BRT on Geary and Van Ness is that in the same way that bacon is deemed by some a “gateway meat,” BRT was intended as a “gateway” treatment to LRV and possibly a subway. Once you have the ROW, then it is always easier to make improvements, you can time BRT trips to “single track” on one lane of the ROW.

    But I also understand that the cost of laying tracks is marginal when compared to the cost of separating the grade, running OH wires and doing the TPS, so it makes sense to lay surface rail with BRT, especially in the Richmond.

    BRT is going to be a bear to implement in the TL, the MTA wants to cut stops to speed up service. The TL is an ideal location for a 4 track subway with express and local tracks that could probably go out underground to J-town or even Masonic before rising to be come BRT. This was part of the Four Corridor plan approved by the voters in Prop B that established the TA.

    Although the Central Subway (completion of another Four Corridor’s project) is not acceptable in its current dumbed down version, going underground is the only option to solve the City’s transportation problems in a way that is appealing enough to compete with the private automobile if for no other reason that our street grids appear to have been designed with the intent to confound efficient surface traversal.


  • jesus h christ. could that thing be any uglier?

  • Peter

    that ‘report’ by WRI is not something to be taken seriously. WRI is bought and paid-for by GM, Dupont, Toyota, and other traditional auto industry corporations. a quick read shows it’s just a propaganda piece. i guess if enough large corporations send you money, then you’re a legitimate operation — not unlike most large news operations (which are not just funded by large corporations, but owned by them).

    there is a reason the most polluted cities in the world are the most polluted cities in the world — they lack real transit. sometimes those cities will put a bunch of big buses on the streets (BRT) and spend lavishly on PR to convince people that buses will make the world a better place. and then the same problems in those cities persist, car ownership skyrockets, and the WRI is off to America to talk about the great ‘success’ they had in some really polluted city somewhere else in the world – a city which is still suffering the toxic effects of extreme pollution.

    still, despite the WRI ‘report’ that ‘confirms’ that buses are ‘the only way to reduce carbon emissions,’ many people cling to old-fashioned notions of reality — we actually require more than just PR puff pieces and fancy youtube videos to convince us that buses are the answer. we’ve seen curitiba and bogota and mexico city and myriad other cities and towns destroyed by hunting buses prowling through their streets.

    san francisco can now look forward to our own array of hunting bendy buses – thanks, in part, to support from all the major ‘green’ organizations in town. i can hardly wait.

    Thank you, GM. Thank you, Toyota. Thank you, Dupont. Thank you, World Resource Institute.

  • LOL @ Marcos: “I want horns here, here and here, and they should all play La Cucaracha

    In the interest of trying not to fan the flames of the argument, I’d like to humbly suggest that BRT just plain ≠ Light Rail, and that’s fine. Just like Light Rail just plain ≠ subways.

    BRT has real, unarguable advantages if you do it cheap and flexible-style. Muni should be establishing dedicated bus only lanes on all of the Rapid lines laid out in the TEP. And I don’t mean diamond lanes like those in the center of Market. The 14, 9, etc need real, enforced exclusive lanes with priority signaling to speed across town. If you can do that for the cost of paint and signal changes, minus the citation revenue from those who illegally use the lanes (which we’ll be enforcing in this fantasy) then you can expand the rapid transit network of SF with buses for a fraction of the cost of LRT.


    If you’re gonna spend hundreds of millions of dollars building a light rail line with no rails, it doesn’t save any money or allow for economical expansion with multi-car trains. The Geary and Van Ness BRT lines are a joke. They seem almost like the dream of someone who wanted to kill transit expansion and improvement once and for all in SF. They will never be converted to rail because they will have cost as much as rail without ever gaining the ridership levels.

  • Note that the study also said that LRT would attract about 20% more riders (but at 100% more cost). I expect this is because LRT is more comfortable, smoother, and lets you read, so the 80% who would also take BRT would prefer the ride on LRT.

    The interesting question is: at what ridership level is LRT cheaper than BRT?

    Labor accounts for 80% of the operating costs of buses, and light rail can save on labor costs by stringing together lots of cars into a train with one driver – if you have enough ridership to fill those long trains.

    This study says that (on this line) light rail “includes higher annual operation and maintenance costs” than BRT. But what ridership do you need before light rail has lower operating costs?

    I would guess that, if you have enough ridership to run trains of three cars or longer, light rail would have significantly lower operating costs, but that is just a guess.

    The next question is when the savings in operating costs is great enough to justify the increased capital cost of light rail.

    Is there any hard data available about these things?

  • I advocate for BRT but only because Muni cannot be trusted with sums of money larger than $1.50. As Mr. Radulovich notes, the T-Third is a disaster. Maybe if Muni/MTA suffered a 100% turnover of staff, I’d revisit the idea.

  • We already have LR capacity that does not expand with multi car (>2) trains. LRVs are a win when they are underground. The only reason to have them above ground is to obviate a transfer, but having them above ground introduces delays in the subways.

    The thinking appeared to be that with incremental large investments, you could move slowly towards a greater capital investment:

    BRT < LRV < Subway.

    What we really need to do is to invest in TBM capacity at the DPW so that we can start a 20 year campaign of subway tunnel boring.


  • danny

    how much would it cost to change the gauge for BART? i know it seems outlandish, but as it was pointed out, the ROW and the tunnel boring is the hardest stuff to do. what if half of BART’s fleet was remodeled with standard gauge axles, and the whole system was re-tracked in a multi-stage blitz? the initial cost would be a lot, but over time, by allowing more train manufacturers to compete for BART contracts, the operational cost could go down when replacing rolling stock.

  • Frederick Gault

    I’ve been to a SPUR discussion about the Geary BRT and it appears to be an elegant solution that can be online quickly at less cost that rail or subway. The passengers would load from the center median so buses wouldn’t be in the way of right turns. There would be a dedicated lane for BRT vehicles with timed lights. The only time traffic interacts with the BRT is at left turns. The studies indicated that this would lower commute times to downtown SF significantly.

  • Muni’s electric power comes from Hetch Hetchy, so Muni’s light rail and trolley buses are 100% fossil fuel free.

    Muni’s electric power is also cheap, since the city owns the dam, but it has to buy diesel power at market rates. Operating costs have to be considered when making transportation choices. BRT proponents argue for the cheaper capital cost (cost to build), but in some cases rail can be cheaper to operate. Operating cost scenarios shouldn’t necessarily assume that diesel will stay cheap. BART did a study last year of whether to electrify the proposed eBART line in eastern Contra Costa County, and found that electrification generated enough operating savings in 20 years to pay for its higher initial cost if the diesel price exceeded $5.50 per gallon; earlier this year, it got to nearly $5 per gallon, before dropping back. If you can ensure that diesel will stay cheap, then by all means build lots of diesel bus lines; if you are interested in protecting against price rises associated with oil depletion, electrified transport makes more sense, because there are many more potential sources of electricity than there are for diesel.

    Ideally, transportation planning in a corridor would not just look at minimizing initial capital costs, or just operating costs, but would seek the solution that optimizes several factors, including initial capital cost, life-cycle operating and maintenance costs, rider benefit (time savings, comfort, accessibility, etc.), environmental benefit, land use and local economic development, and future capacity needs. Neither the transportation planning at MTA nor the SFCTA’s Geary and Van Ness corridor plans have done this. The SFCTA studies over-emphasize initial capital cost, while recent MTA rail projects (3rd Street and the planned Central Subway) manage to assure both high capital costs and high operating costs while delivering little additional capacity or rider benefit.

    Priority one is to figure out how to manage MTA’s existing rail system better. It is frustrating to watch packed two-car trains moving through the Van Ness station at rush hour, while the platforms could easily accommodate four to five car trains. Using low-floor vehicles and proof of payment could speed up boarding tremendously, and managing streets better (what’s up with all the stop signs on Church Street?) could get trains moving much faster.

    Livable City got MTA to promise to commission a plan for how to better manage the rail system (including resolving the high-floor/low-floor question) as a follow-on of the TEP, but I haven’t heard yet when they intend to begin the study.

    As we figure out how to manage our rail system better, we should also figure out where it makes sense to add or extend rail lines. I an alternate universe where San Francisco plans transportation intelligently, Geary would probably be a rail corridor, and Third Street might have been BRT.

  • I neglected to add that labor costs are typically the largest transit operating cost, and energy is number two. Rail’s operating cost savings comes with large passenger volumes; trains can be much longer than buses, meaning that one rail operator can potentially carry many more passengers than a bus driver can.

    MTA manages to negate that potential operating cost advantage by running no trains longer than two cars. Imagine if Muni ran three-car rush-hour trains on the N-Judah; operating costs would increase only slightly (still one driver, but slightly higher energy and maintenance cost), but capacity would increase by 50%.

    MTA also made bad choices about rolling stock; the Bredas are among the heaviest, and most expensive, light rail vehicles around. Lightweight low-floor cars would be cheaper to buy, and cost less to operate.

  • theo

    What I don’t understand is the MTA’s inability to pull the trigger and switch over to electronic fare collection with Translink.

    The potential savings are enormous.

    Say each rush hour N Judah train gets 300 boardings.
    About 1/5 of riders currently pay cash, so that’s 60 cash boardings.
    Muni consultants have estimated that half of all cash users will switch to TransLink, so that’s 30 cash boardings eliminated.
    Each cash boarding takes about 15 seconds.

    That’s 7.5 minutes time savings per run, which is more than even the TEP promises!

    This would be a huge boost to speed, reliability (less bunching) and frequency (because the same trains could be run more often).

    Why is it taking so #$#% long?

  • It looks like MUNI Planning needs a Rescue MUNI Planning.

    Also not acknowledged is that in NYC, my understanding is that the Subway has been shown to increase business productivity by a measurable multiplier.

    It is truly astounding to see enviros opposing making a significant investment into the kind of rapid transit electric traction infrastructure that lasts for centuries.

    Had these ideas been studied in the past instead of rejected out of hand, we might have been better positioned to take advantage of stimulus spending to build the quality level of transit investment that we need.


  • Next time you ride the 14, 38 or the 9, take a stopwatch and time the delays from 2 sources: traffic/signal delay and boarding delay. You usually get more delay from boardings. Lowering boarding delay is such an easy and cost-effective measure: Space stops no closer than every 1/4 mile and switch to Proof-of-Payment, double door boarding. Faster operating speeds means less drivers are needed to maintain headways. Use that cost savings to hire fare inspectors for the bus network.

    BRT does not solve the 38’s real problem: overcrowding. Rail is needed and warranted on America’s busiest bus corridor outside of NYC. Subway is needed on Inner-Geary, especially with the number of pedestrians and the narrow streets.

  • I haven’t read any of the comments yet, but Matthew, you’re just looking for a fight.

  • When do we get a Metro? Not just for San Francisco but for Oakland and Berkeley. We have a dense city, compact and walkable, why should it take me 40 minutes to go three miles? Imagine the implications of a real metro system. I’d gladly pay much more for a system that Muni couldn’t mess up. Though perhaps that is a hope too far.

  • Peter

    i somehow missed what seems like an entertaining read from this past summer (2008). The uber-serious ‘Transport Reviews’ publication, in its 4th edition of 2008 (the same one with Pucher’s “Cycling For Everyone” article), also has an article entitled, “Bus rapid transit: is Transmilenio a miracle cure?


    Successful mass transit solutions are rare in poor cities. When they appear they are lauded across the globe and too often copied uncritically. The latest exemplar of such best practice is the ‘Transmilenio’ rapid bus system in Bogota. The article describes its main characteristics and applauds the improvements that it has already brought to urban transport in Bogota. Naturally, the system is not without its flaws and these need to be drawn to the attention of those who might copy the Bogota example. This is particularly important at the present time when the jewel of Bogota has come under surprisingly strong local criticism over its cost, its ownership structure, its decreasing effectiveness and, fundamentally, because it has failed to solve the transport chaos of Bogota. There is a real danger that ‘Transmilenio’ will stagnate as its popularity declines and as demands for a metro increase. Given the strengths of the system that would be something of a disaster and, most certainly, not in the interests of the poor.

    I couldn’t find a free copy online anywhere — not that i expect anything could stop the BRT train from running over SF next, but I am curious to know more about what’s popping in the city where the transportation is strong, the buses are good looking, and all the bike paths are above average.

    Or something.

  • “The big problem with San Francisco is that we build incredibly expensive light rail, like the Townsend 3rd, and we run rail like a bus. We’re not being smart about our investments.”

    Good to see, I’m not the only one that notices.

    “[I]sn’t it mathematically impossible to time signals on a two way street so that transit vehicles never have to stop or to run at speeds that would compensate for those delays?”


    “It is truly astounding to see enviros opposing making a significant investment into the kind of rapid transit electric traction infrastructure that lasts for centuries. Had these ideas been studied in the past instead of rejected out of hand, we might have been better positioned to take advantage of stimulus spending to build the quality level of transit investment that we need.”

    Marco they have been studied. They’ve just been ignored. The trifecta of light-rail lobbyists, ambitious/ribbon-cutting/term-limited politicians, and misguided environmentalists have led to the current state of affairs.

  • bikerider

    Really no surprise that LRT compares poorly against BRT. But does this mean LRT is obsolete, or merely that Americans do an extremely poor job in implementing LRT projects?

    Compare almost any LRT project on the West coast to the standard European tram line. Our capital costs are several orders of magnitude greater. The difference in operating cost is absurd too. Unfortunately, I fear there is little chance that Muni (or any other US transit agency) will develop a level of competence that European transit riders take for granted.

  • It isn’t impossible to time a two-way street where cars in both directions get green the whole corridor (see this). But it will likely adversely affect a bus that needs to stop every block or so because the bus will get “behind” the vehicle cohort that the signal was designed for.

    But this is all moot. Traffic signal priority systems can calculate when a bus will be arriving to an intersection and adjust the signal appropriately for a two-way transit corridor.

  • anonymouse

    The initial build of BRT might be cheaper, but is that really what matters? Pavement is much less durable than rails, and buses have to be replaced 2-3 times more often than trains. And of course, there’s the capacity difference between a single 60 foot bus and a two-car light rail train (or three or four car, if Muni ran the subway competently). And to all the proponents of more subways, I would say: think hard about total door to door travel time. Having a subway means less frequent stops, so more walking to the stop, and just the fact that it’s below grade imposes a time penalty on either end to get from street level to the train and vice versa. SF is a small city, and it might not make sense given its size to have subways, at least outside of the CBD.

  • The whole point of a 4 line mainline serving critical mass corridors, which would be Van Ness and Geary, is to have both local and express stations and that is the win, where once you’re in the system, you’re shortly on the express line and getting where you’re going faster.

    Local stops between Midtown and 14th Street are about .25 miles apart while express stops are closer to 1.5 miles.

    Unless transit is made attractive as part of a regional high speed network, Transit Oriented Development is makeup on the pig of greater densities, more expensive luxury condos, more cars fighting for street spaces and higher developer profit.


  • Tom Radulovich for MTA director! Actually, no, we’d miss him too much at Livable City; who would be the voice of reason and policy improvement?

  • Minor nit – energy is “fungible” (my new word of the day). So to say that since MUNI uses energy from Hetch Hetchy and is thus 100% fossil fuel free is not an accurate statement. That energy – were it not used by MUNI – could be used elsewhere. The city does not produce more energy than it needs by renewable means.

    This doesn’t detract from anything else Tom says.

  • Regarding the “time penalty” on subways:

    1) There is no penalty in comparison to street-level because the travel speed is far superior, especially in comparison to the type of street-level transit that actually get passengers closest to destination.
    2) Subways provide the opportunity of constructing portals closer to desired destinations if not WITHIN the destination, which leads to my last point:
    3) The distance from platform to surface (and thereby time it takes to traverse) is primarily a design issue.

    SF is a small city, and it might not make sense given its size to have subways, at least outside of the CBD.

    I fully admit I don’t know enough about San Francisco land use, communities or travel patterns to register a definitive opinion on whether subways are needed or where. But I do know that the determination as to appropriate mode isn’t solely limited to the population of a city, especially not one that is part of a much larger metropolis.

    By your logic, Boston is a small city (smaller than San Francisco by the way) and doesn’t need it’s three (should be four) subway lines.

    Unless transit is made attractive as part of a regional high speed network, Transit Oriented Development is makeup on the pig of greater densities, more expensive luxury condos, more cars fighting for street spaces and higher developer profit.

    And the church said, “Amen.”

  • buses: least expensive. fewest riders. slowest(especially in san francisco, hello 30 stockton).
    brt: more expensive than a bus, attracting more riders than a bus. a little faster.
    lrt: more expensive than brt, less expensive than a subway, attracting more riders than brt and less than a subway. faster still, if not operating in traffic.
    subway: the most expensive, the most riders. the fastest.

    when transit agencies are striving for half-assed solutions to our needs, it’s not a good thing. brt is nothing more than a fancy, expensive bus with a more aerodynamic(ugly) exterior and maybe more comfortable seats. it’s a band-aid. yeah, muni can’t be trusted with money, but could you tell me who could?

    what happened to *demanding* real solutions to real problems? other options are more expensive and work better for people who use it.

    regarding the environmental aspect – if you don’t want to pollute, ride a bike or better yet, walk.

  • bikerider

    Subways are no faster than LRT. Nor are subways/LRT faster than BRT.

    What makes a subway “fast” is the long station spacing. An LRT system with subway-like station spacing (sometimes called ‘stadtbahn’) has speed profile comparable to standard underground metro (there are any number of real-world examples of this).

    Again, the only difference is cost. Underground stations are hideously expensive to build (which is the reason for having fewer of them, spread out further). LRT and BRT do not have this problem.

  • Where there is room in San Francisco for a Stadbahn? Is there a rail ROW we’re not aware of out there? If you’re going to have surface LRT that can go as fast, you’re going to need grade separation to make it competitive with a subway option which means elevated. In an urban setting like San Francisco, you’ll need higher capacity transit if you’re going to make it so fast that people want to take it. 10 car BART trains will not fit on the surface because they would take up too many blocks. That is why LRT in Portland only uses two car trains, because when it gets downtown the blocks are too short and it would be blocking streets. If they built a subway they could accommodate more people, longer trains and get greater efficiency out of the system. We should have three car trains at peak on the N. It would make the cost per passenger mile go down.

    But that is only really an option for a few corridors like Geary. Put a Subway on Geary and I guaranty 100,000 riders. 100,000 riders on a single corridor? That is a great investment over the long term and I think somewhat conservative, for operations, for the environment and for the people of SF who cram in the 38 every day and those who don’t because its too crammed. The Bay Bridge carries 270,000 cars a day and we’re paying over $6B for it. Let’s get a TBM like a poster above said and start building it from the Richmond in. When it gets there we can train it on another corridor. It can’t cost more than $2B for 6 miles of Subway. If we’re gonna do something, let’s do it right. Let’s not half ass just because of cost, because it will cost us more in the long run in negative externalities. Seems like a perfect stimulus package to me.

  • bikerider

    While technically not a ‘stadtbahn’, the new tram lines in Paris carry over 100,000 trips per day, per corridor. And squeezing tram lines in Paris is WAY harder than Geary (or any other location in SF).

    See Plenty of other examples in Germany too, if you bother to do the research.

    To reiterate: the decision point of BRT vs. LRT vs. underground is driven mainly by expected passenger loads, not vehicle speed. If your local transit agency can’t make buses run as fast LRT, then it is doing something wrong. And if it is spending $1.5 billion building underground metro that carries fewer than ~10k trips/hr, then it is doing something VERY wrong.

  • I’ve got serious problems with the “first day ridership” metric for transit projects. These are systems that will last for centuries and will bring centuries of efficiency to San Francisco’s economy.

    The argument is that MUNI riders need to suck it up and accept slow travel times, offloading the costs of MUNI’s inability to plan for rapid transit into the wallets of hundreds of thousands, either by time wasted in slow, crowded, yucky buses or in the cost of a car to blow by the slow buses.

    Any system that is constrained by San Francisco’s street grids is going to underperform a subway not because of stop spacing, that problem has been solved 100 years ago in NYC with a 4 track mainline, rather by having to deal with surface traffic.

    Again, even if there are BRT/TPS treatments on a major corridor that grant transit priority, that win will only be at the cost of deprioritizing adjacent transit which must wait for the mainline.

    And since MUNI is welded to this radial line notion, which prioritizes getting people to work downtown over facilitating San Franciscans’ conducting our errands quickly and efficiently by transit, it would make sense to expand on Josh’s proposal for a light rail 22 bus by adding the 44 and 33 to the mix for a peripheral rapid transit network. Perhaps the local stops could be served by surface vehicles while the subway only stopped at key points where surface riders could transfer below ground.


  • The Budapest Tram carrries over 200,000 riders a day on the 4/6 line. It’s not as fast as a subway because it has street lights to contend with What’s your point? Throwing out ridership numbers doesn’t tell us anything. The Paris surface light rail is still in street and as Marcos said it’s an issue of street grids and stop lights and speeds possible in an urban environment. If you got LRVs going 60mph in the median of Geary I don’t think there would be many happy community members. Underground however is another story.

    And Marcos, you’re right…it is a long term investment and should be viewed as such. Right now the Fed rate is at 0%. I don’t think we can get much better than that for borrowing money to build long lasting transit.

  • bikerider

    1. The 4/6 tram is not as “fast” (and I use that term loosely) because it has stops every .25 miles. By comparison, Budapest metro stops every ~0.5 miles (typical of urban metro).

    2. Even if a subway were built along Geary, it would be extremely counterproductive to operate at 60mph. That would entail some combination of: miles-long station spacing, high energy/maintenance costs, and/or poor operating frequencies.

    3. For the speed and frequency we are talking about, traffic signals are simply a non-issue. The nice thing about the exclusive lane is that the transit operator can schedule for their trains/buses to arrive at a traffic signal with 99% certainty, which allows the signals to be timed for arrival of each passing train/bus (in the 1% case where it is a few seconds late, the TPS and computer-synchronization can hold the light a bit longer).

    To use real-world example: AC Transit plans to run their BRT along Telegraph at 2:30 intervals in both directions. Their buses will have exclusive access to traffic signals, without impacting heavily-traveled cross-streets like Ashby.

  • “The passengers would load from the center median so buses wouldn’t be in the way of right turns.”

    Don’t tell me the median option is now the preferred!…

    Sounds like a great idea doesn’t it? BRT stations in the middle of the road, out of the way of cars and out of the way of riders :-/

  • bikerider:

    It’s not possible with a frequently-running LRT to perfectly synchronize lights (giving full priority to trains) on a major metropolitan street to where they’re never or rarely stopped unless the transportation agency is willing to endure gridlock on both cross streets and possibly adjacent traffic as well.

    In the case of LA, if it were possible to do such without leading to total gridlock, we would have figured it out a long time ago with the Blue Line in Downtown LA, where we have exactly what you’re advocating: 3-car LRT serving stations with Metro-like spacing (every 0.75 mile). Here even the adjacent traffic is impacted because often the left turn lanes/lights aren’t long enough and the traffic hangs-over into the through lanes. Instead, some priority is given to trains but there still will be at least two-three delays in a short 2.5 mile street-running section during rush hour, sometimes as long as 180 seconds.

    It is not possible in practice, and even the best system takes into account the fact that the transit mode will have to interact with humans on a busy thoroughfare. I’m talking about passengers that take a bit longer to load/cross the street, spill over traffic in the intersections, etc.

    And the operating speeds of LRT even with prioritization on the surface is capped at 35 mph (it’s state law). That’s not the case below the surface (unlimited). Even with stations every 0.5 mile grade separated trains can typically max out at 42-45 mph comfortably. With stations spaced every mile, you’re looking at a comfortable max speed of 64 mph.

    Surface level capacity is limited. It is needed not just for vehicular movements, but also to maintain reliable connecting bus service, street amenities and pedestrian movements. People always seem to forget that there needs to be enough time in the cycle to allow pedestrians to cross safely. There’s only so much the surface/traffic cycle can give and allocating some of it to the train means less for not just for traffic lanes and left turn lanes, but also for bike lanes, curbed parking (which creates a buffer for pedestrians), sidewalks widths and pedestrians.

    And to piggy-back on what The Overhead Wire said, ridership numbers tell you nothing. You can’t just pull up a transit system in one city, plop it down in another and expect it to perform the same. It’s much more complicated than that.

    Again, I wouldn’t go so far as to say I know what and where the subway lines need to be. But I have been dealing with several of these general discussions about LRT vs. grade separated transit and find many of the assumptions/theories made by those who argue against grade separated transit to be (for lack of a better term) simplistic and seriously faulted. For example, and it’s just one example – I could have picked others, there’s no such thing as “99% certainty” for any mode of street-level transit even with a dedicated right-of-way. Come down to LA and I’ll show you. Here we define “on-time” for our street-running rail as “within 2 mins before and after scheduled arrival.” A 2 min delay in some other countries subway systems is consider totally unacceptable.

    There’s a good body of academic work, some of which I’m sure anyone could find on the web, about capacity limitations of street-level transit, both for the rail transit mode and the transportation system as a whole. The rest of the world did figure this out over a century ago.

  • bikerider

    LAMTA is one of the most retarded transit agencies when it comes to rail transit. While that perhaps makes it a good analog for Muni, we should not hold it up as a shining example of what industry best-practices in transit operations has to offer. Surely we can do better than “within 2 minutes” as being an on-time arrival for LRT.

    Your observation about LRT causing delay on adjacent (car) traffic in LA is no doubt valid, and a reflection of mushy-headed LOS-oriented traffic planning in the US. By contrast, the approach with the European tram systems is to not worry so much about the automobile LOS impacts. In many cases, the car traffic is eliminated entirely, with congested downtown areas pedestrianized as part of the tram project. Granted, that approach is not so politically feasible in the US. But let’s be clear then that you are suggesting subways as a political, not a technical solution; i.e. spending many billions of scarce transit funds to move trains underground just so cars don’t overflow a left-turn lane waiting at a light.

    A few other comments:
    1. Spending billions to increase max speed from 35mph (or even 25mph) to ~45mph is not cost-effective. SF is only 7 miles in any one direction, I’ll leave it to you to do the math and compute maximum theoretical time savings. Then compare your time savings against: 1. time to schlep down an elevator/escalator to underground platform; 2. longer walk to subway stop

    2. Indeed, there is huge amount of work on transit design and capacity, and yes this stuff was figured out a long time ago. So why is it that Muni (and LAMTA) planners don’t follow industry-best practices? The Bay Area actually spends just as much as our European counterparts on new transit projects, but with very little to show for it.

  • Bikerider,

    1. Funds are not scarce when interest rates are very low.

    2. Current BRT proposals will spend hundreds of millions to decrease 48th to Embarcadero travel times via Geary by what, 15%?

    3. If travel times could be reduced by 50% with a four track express/local configuration for around $2b+, probably less with falling prices for construction inputs, then stops would be slightly more spaced out than now on Geary, but once on the express, you’d be downtown in 15 minutes. Ciompare that with the 38L in the middle of the day with no traffic under conditions as similar to BRT as you’ll see in the real world.

    4. The cost savings over time from running longer trains underground faster will pay for the cost of the system over its useful lifetime.

    5. The environmental savings of electric traction advance other policy goals

    6. More reliable subsurface transit gets people out of their cars, off the roads and out of the way of me and my bike.

    7. Reaganoid arguments about how government is too incompetent to solve our problems don’t play here.

    8. These problems were all solved at the end of the 19th century, it is just that we are too damn cheap to have learned those lessons.


  • Spending billions to increase max speed from 35mph (or even 25mph) to ~45mph is not cost-effective.

    C’mon now. People are smarter than that. The realize you’re not just increasing max speeds, you’re increasing travel times because you’re eliminating the need to wait at traffic lights.

    What you advocate, street-level street-running with max speed of 35, averages around 15-17 traveled miles per hour, and that’s with stations spaced every mile. Space them any closer to take advantage of your stated street-level advantage of getting closer to the destination (which again, I don’t agree with by the way as stated in #27) and that number plummets even further.

    For comparison, grade separated rail is consistently 28-33 mph, with express service in the territory of 38 mph.

    Indeed the primary reason many newer LRT systems on the West Coast, AVERAGE higher traveled miles per hour is because significant portions of the routes operate like a Metro: grade separated, often in freeway medians, adjacent to highways or underground in urban centers. It’s a hybrid – not all street-running like you advocate.

    The many other drawbacks of surface level rail raised in previous posts, which I don’t find the need to repeat, are just as, if not more, overwhelmingly in favor of grade separated transit as the speed advantage.

    Regarding your use of what I like to call “generic out-of-town examples.” Typically I respond by requesting a person not generically use an example, (you use “European tram systems”), but rather give us specific cities and transit systems.

    Specifically with Europe, often they don’t because they quickly find out that almost all of the major cities in Europe that they’re thinking of, which have trams, like MUNI, also have extensive grade separated systems, like BART. They realize the trams are not the REPLACEMENT of the Metro they are the compliment, and the job of distance travel (rapid transit) is handled by systems that are completely (or primarily) grade separated. The few exceptions I’ve seen are all in countries that experienced severe financial hardship during the Soviet Empire and/or prior.

    We didn’t experience severe financial hardship in San Francisco, LA, or Seattle (cities with densities and populations that on paper are comparable to east coast cities with extensive Metro systems) yet we’ve failed to make the necessary investment (especially in LA and Seattle). Past and current incompetence and inefficient investment is not persuasive argument for future direction.

    In #27 I also addressed your arguments regarding your limited focus on the 7 miles of San Francisco, as though the 2.5 million other people who make up the urban area don’t exist.

  • I forgot to mention that the 15-17 traveled miles per hour for street-level LRT is WITH 10% signal priority, the “industry standard” by the way.

    WITHOUT 10% signal priority, skip-cycle, and all the other tricks that traffic departments have had to integrate, again the traveled mile per hour number plummets.

  • Nathanael Nerode

    “And the operating speeds of LRT even with prioritization on the surface is capped at 35 mph (it’s state law). ”

    What the HELL is wrong with California? That is an absolutely ludicrous state law. The operating speed of LRT should be limited to the same speed as the auto traffic on the same road, approximately; if it runs down a 55mph road, there is absolutely no sane reason to limit it to 35 mph. (Not that this is relevant to downtown San Francisco, where I think all the roads have 30 mph speed limits.)

  • I have not much time, but I’ve got many useful things
    here, love it!


The Future of Van Ness Avenue is a Full-Feature BRT Route

Van Ness BRT at City Hall, Alternative 5: center-center median option. With overwhelming approval for the Proposition K half-cent transportation sales tax in 2003, San Franciscans  signaled they not only wanted to maintain a state of good repair and operational solvency for their transit system, they were willing to dedicate more than 25 percent of […]

San Leandro City Council Meeting: East Bay BRT

From EBBC: The final votes for East Bay Bus Rapid Transit come up July 16, 7:00pm at San Leandro City Council and July 17, 5:30pm at Oakland City Council. We need supporters in attendance ready to boost councilmembers who want to vote yes, but who need our backup. Bus Rapid Transit (BRT) is a system that takes […]

Hampered by Tunnels, Center BRT Lanes on Geary Limited to the Richmond

Correction 12/17: The next community meeting on Geary BRT is tonight, Tuesday, at 5:30 p.m. at the Main Public Library. The latest iteration of the plan for bus rapid transit on Geary Boulevard includes center-running bus lanes only on the Richmond District segment between Arguello Boulevard and 27th Avenue — about a quarter of the street’s […]