Wednesday, February 27, 2008

Ending the BRT Bait-and-Switch: How Rapid?

For years now some transit advocates have been touting systems under the heading of "Bus Rapid Transit," arguing that it can bring many of the benefits of rail rapid transit at a significantly lower cost.

A lot of transit people are skeptical of BRT, with good reason. BRT is often used, quite consciously, by people opposed to transit (or opposed to funding transit with taxpayer money). Under some definitions, BRT can include everything from rubber-tired underground or elevated metro systems such as those in Paris and Montreal to the same ol' bus with a new paint job. Here is a list of BRT characteristics from Wikipedia:

  • Bus only, grade-separated (or at-grade exclusive) right-of-way

  • Comprehensive Coverage

  • Serves a diverse market with high-frequency all day service

  • Bus priority / Bus lanes

  • Vehicles with Tram-like characteristics

  • A specific image with a Brand name

  • Off-bus fare collection

  • Level boarding

That scale can be abused by unscrupulous people to divide and diffuse support for rail. Often it seems that the agency in question spent a lot more time and money on branding than on anything else. Image is important, but it's not everything.

In particular, the "BRT" pilot project for New York City is particularly disappointing, where each corridor was proposed to get a minor, heterogeneous collection of improvements. The BRT documents have been MIA from the MTA's website for some time now, and my hope is that the whole plan is getting a hard look from people at the DOT, and will emerge much better than before.

So here's my proposal to cut down on the "BRT" bait-and-switch. Let's focus on the word "Rapid." Rapid should mean something, and that something is "fast." Fast may mean different things to different people, but I think for every project, the affected people should agree on how fast is fast, and anything below that is just not Bus Rapid Transit.

There's often no concrete prediction for time improvement (such as a minimum speed, or areduction in trip time). In the absence of a specific number, focus on a specific feature. It may be that there's a corridor out there that could get major increases in speed through level boarding and signal prioritization alone, but the primary factor is right-of-way, and the single most important feature is physical separation of the right-of-way. It's hard to be rapid when you're stuck behind double-parked cars.

So the next time someone says "Bus Rapid Transit" to you, ask them, "How rapid?" If they can't give you a number then say, very sweetly, "Well of course the right-of-way will be physically separated for most of the route, right?" If they can't even promise that, then tell them to come back when they've got the numbers. Otherwise, they can call it Bus Slow Transit and abbreviate it BST, pronounced "bust."


Pantograph Trolleypole said...

I was annoyed when I saw a streetsblog post that showed the Capital costs for BRT versus rail. Their number stated $13 million per mile but I know for a fact that true BRT costs a lot more. The lowest recent costs have been $22 million for LA's Orange Line and $23 for the Euclid Corridor in Cleveland. High costs include the Hartford Busway which will cost $55 million per mile. These are true BRT lines which 100% dedicated guideway, not half and half like Eugene and other cities.

Anonymous said...

San Francisco BRT projects in planning are estimated to be closer to $40-50 million/mile, but those include pretty substantial streetscape improvements as well.

You're right that there is definitely confusion about what "BRT" actually means. The term has been applied to a lot of watered down projects that would more accurately be referred to as "limited bus with a little signal priority" or some such.

Personally, I reserve the term "BRT" only for truly thorough implementations, in which all or most of the route has dedicated ROW, along with signal priority, and rail-like infrastructure: stations, ticket machines, real-time information on bus arrival. It should feel just like a modern light rail line, except the vehicle you step onto is a bus. If a project is any more watered down that, I might call it "BRT-like", "updated", "improved", or "offering some BRT features."

Bus improvements are basically a continuous spectrum, but really only the very upper end of that spectrum should be given the term "BRT."

Cap'n Transit said...

Thanks for your comments! Pan, the Streetsblog and Tri-State bloggers are good people. You should post that comment on both blogs.

Transbay, I see your point, but I'd like to have a bit more flexibility in the definition of BRT. I don't think that any of the goals of BRT require it to feel any particular way, other than Rapid.

For improvements that don't qualify as Rapid, I like the term "Quality Bus."

I'll also point out, as has been mentioned on Streetsblog, that it's not always necessary to improve the image of bus service. There are places where buses have prestige, and all that's necessary is to make them more effective.

Pantograph Trolleypole said...

Yeah I love Streetsblog, they do great work. But the whole BRT thing really confuses everyone. Like in Oakland and LA they have Rapid Lines. Well they just took the L for limited off the name and stuck an R there. Then people outside of the region start calling them BRT. They still run in traffic, just have a fun paint job, which is what you're saying. In Houston they are going to call their limiteds something completely different. Perhaps someone (APTA) should set a standard as to what you can call your service so people don't get confused and think they are getting one thing when it's really just an L.

Alon Levy said...

This is the first time I see anyone refer to the systems in Paris and Mexico City as bus rapid transit. The rubber-tired Metro lines involve cars articulated to form long trains, and guideways made of rail and grade separated from all other traffic. That's not BRT; that's heavy rail.

As for the cost... $55 million per mile is pretty low. Second Avenue Line is officially supposed to cost $16 billion and be eight miles long.

(Of course, you can radically shave costs by running trains in highway medians, but you still have to connect with other rail transit. Every destination that's close to a highway is built to automobile scale, so telling people to get off the highway train and start walking is unrealistic.)

Pantograph Trolleypole said...

The second avenue line is also a subway under the tall buildings of New York City. $55 is what you'll pay for light rail if you do it right.

Anonymous said...

We need to agree on the definition of "fast" (or "rapid"); it's a key concept.

I propose we define "rapid" to mean "faster than you can get there driving your car". Any system that can't get you there faster than driving your car shouldn't be able to call itself "rapid".

This avoids bogging down in m.p.h., since "rapid" can mean different speeds, say, between cities than it does in within neighborhoods.

Alon Levy said...

It's a subway under a street; the only part that runs under buildings is in the last stage of the project, which even the official estimates say won't be built until 2020.

Cap'n Transit said...

Alon, people don't actually refer to the Paris Metro as "BRT," but they say, "Well, you know the Paris Metro runs on tires," leaving you to think that the key is making your bus look as much like a train as possible. You're absolutely right that it's bullshit, but they say it. I wish I could find a quote.

J, here in New York the subways are faster than cars. There are a lot of places, though, where the trains are not faster, but still at a competitive speed. I think "at least as fast as you could get there by car" is good.

I think busways that have at-grade crossings with stop lights, but also signal prioritization and physical separation, can't really go faster than cars (they still have to board passengers), but are stil competitive. What do you all think, are they rapid?

Alon Levy said...

No. The chief defining features of rapid transit are that it's fully grade-separated from other traffic, and that it's based on cars that can be articulated to form long trains. Articulated buses don't go much above 20 meters; metro trains can easily clear 200.

Speed is actually not that important. In Singapore, cars are considerably faster than rapid transit; many arterial roads are congestion-free even at rush hour, enabling average speeds of 60 or 70 km/h.

Rather, what distinguishes rapid transit is high capacity. BRT proponents love to talk about the high capacity of the Lincoln Tunnel's XBL. But when you don't have to stop on the way, the capacity is always higher; the XBL works only because it feeds directly into a gigantic terminal. In a more common setup, BRT capacity is in the high thousands of passengers per hour. Meanwhile, the RER is on the far side of 50,000/hour on its busiest line.

Cap'n Transit said...

Okay, Alon, I'm not trying to give you a hard time, but is the chief defining feature of rapid transit grade separation or high capacity?

If it's grade separation, then several lines in the Chicago area (the Douglas, Ravenswood and South Shore lines come to mind) are only partially "rapid." Many of the popular light rail lines, like the Tramway des Maréchaux, are not rapid.

I agree that BRT isn't likely to greatly surpass the capacity of the Lincoln Tunnel XBL, so it's a good maximum to work with. By my calculations it carries 16,500 people per hour. So is 15,000 per hour enough to count as "rapid," or even 10,000?

I think to answer that question we need to go back to our goals. What do we want out of BRT? A lot of the goals are dependent on "getting people out of their cars," which requires providing a quality of service that's better than driving.

As long as it's Rapid, BRT can provide a superior quality of service to driving. I think that's enough to serve non-drivers and have significant impacts on land use.

Capacity is what allows the service to scale, so that it can get more people out of their cars. It's a good question whether BRT can get enough people out of their cars to affect the larger issues, like adding efficiency, reducing carnage and pollution and curbing the demand for sprawl. I think probably not enough, but that doesn't mean it doesn't count as "rapid transit."

Alon Levy said...

My understanding of the L is that some of the lines run in highway medians, but are still separated from all other traffic. It's permissible for heavy rail to run at grade as long as it doesn't have to cross in front of cars. It's even permissible to have some at-grade crossings - NYCT had some into the 1960s - as long as they're traditional railroad crossings rather than stoplights, shunted away from the busiest lines, and too infrequent to constrain train length.

Of course, light rail has at-grade crossings. It's glorified streetcar service, or sometimes a hybrid of streetcars and metros. Streetcars wait at stoplights just like automobiles; so does light rail. This forces shorter train lengths, on the order of 30 or 40 meters, and limits capacity to about 20,000.

As for what we want out of BRT, I'd say we want it to provide service to areas too lightly populated for light rail, and we want light rail to provide service to areas too lightly populated for rapid transit.

Cap'n Transit said...
This comment has been removed by the author.
Cap'n Transit said...

Hm, lemme fix these links...

Alon, the Chicago lines I'm talking about don't run in highway medians, but in separated ROWs at grade, like this - with traditional RR crossings. The South Shore line runs like this in Michigan City. I don't know how much of an effect either of those alignments have on speed.

When I talk about goals, I'm not talking about what we want from BRT, but what we want for a particular corridor. Then we judge whether BRT will fit those goals. I take it you don't think that either BRT or light rail are Rapid. So why continue to use the term "BRT"? Call it BST, or Quality Bus Service.

Alon Levy said...

Hey, I didn't invent the term. I think you can talk about BRT as opposed to ordinary buses if the stations are more widely spaced, but BRT is still just glorified bus service in the same way light rail is a glorified streetcar network.

For a particular corridor, it depends on its nature. Some corridors are supposed to funnel people from a large area into one point, for example the bus services in Eastern Queens that converge at Jamaica Center. In cities with subway systems with large interstations, such as Singapore and Moscow, there's also bus service that gets people to subway stations in denser areas. Some corridors are specialized, like those connecting cities to airports. Others are meant to increase connectivity along one important street, like Flatbush or 125th or 34th.

Buses are good for the first two categories. When the streets are empty, they're even good for the last one. Singapore has a (small) congestion fee, fairly high gas taxes, onerous car taxes, and an infrastructure that would make Robert Moses proud, so its buses are reasonably rapid, and get people precisely to where they want to go.

In more congested areas, I'd prescribe light rail for the last category, airtrains for the third, and buses for the first whenever population density doesn't justify rapid transit.

(P.S. although a car-oriented infrastructure can make buses cruise given anti-car regulations, it also makes it hard to run good rapid transit. Singapore has no good corridors for rapid transit, since the major streets are built for cars; its retail and office areas come in clusters rather than corridors except for a short stretch in the center of the city.)