Pros and Cons of Center- vs. Curb-Running Bus Rapid Transit


The City of Boston has created a pretty nice one-sheeter laying out the advantages, disadvantages, and design principles of running exclusive bus lanes in the center of a road vs. along the curb. As the Portland region considers BRT for future rapid transit lines (Powell/Division, Clackamas to Washington Square, and the SW Corridor are all under consideration for BRT), it is worth thinking about these principles and trade-offs, so take a look!

Center-running bus lanes in The Hague

In general, a center-running alignment is preferable if we want a high level of service. It eliminates conflicts with right-turning vehicles and bicycles, generally gives exclusive signal phasing for transit vehicles, and it breaks up wide streets in a way that can dramatically improve pedestrian crossings. There is a reason why virtually all transit (both trams and buses) in the Netherlands runs in the center of the street, and I can report from experience that it works very well and breaks up wide streets very nicely. Of course, we don’t have to look that far for examples. Our very own Yellow Line on Interstate and Blue Line on E Burnside follow the same principle, although I would argue they could be retro-fitted to be a lot more attractive.


Center-running transit can be quite attractive

Unfortunately, because Bus Rapid Transit is often chosen over Light Rail Transit as a way to save the most money possible on a project, it is all too often relegated to the outside lanes where performance is not as high. Both RapidRide in Seattle and Swift north of Seattle are examples of BRT lines that run in the curbside lanes, either in mixed traffic or in semi-exclusive Business Access Transit (BAT) lanes that are shared with right-turning and parking vehicles. While BAT lanes are better than nothing, they can often be just as clogged with cars as regular lanes, are difficult to enforce, and are often compromised (as we see through Interbay in Seattle) by making them peak-only.

As far as I can tell, there are two main reasons to choose curb-running over center-running lanes. First, if there is not enough right-of-way to fit in the center platforms. Second, if there is a political unwillingness to completely take away car travel lanes. The first reason doesn’t really apply to the roads being considered for BRT in Portland. Powell Blvd, even in inner SE Portland, has plenty of room to make the center lanes bus-only and still install median platforms. All we have to do is remove the center turn lane and the planted median. Same goes for Barbur Blvd in SW Portland, which has 3 car lanes in each direction and a small painted median.

The second reason is what will determine whether Portland ends up with high-quality BRT or not. People will object to removing any car capacity on Powell or Barbur, insisting they are critical auto corridors that should not be significantly altered. Curbside BAT lanes (probably peak-only) will be proposed as a way to “balance” the needs of cars and transit. ODOT continues to insist that Barbur needs its current capacity to act as an “overflow valve” for I-5 in case of accidents or congestion. These arguments fail to convince after the experience with N Interstate Ave. That was a road that was considered an important auto corridor and an overflow valve for the northern section of I-5, and yet we reduced its car capacity in favor of building MAX and life goes on. People have adjusted to the new reality and I am unaware of any major problems with the new configuration.

In the case of Powell, a reduction of lanes and a corresponding reduction in speeds would do wonders to improve the street’s safety, walkability, and livability. We should also remember that any BRT would be able to get off Powell somewhere around SE 17th to head up to the new Willamette transit bridge, so extra capacity would be available to deal with the Ross Island Bridge bottleneck. In the case of Barbur, that road is over-capacity anyway and should not be kept that way simply to absorb the occasional spillover from the freeway. Why do we have a 6-lane freeway and a 6-lane highway running parallel anyway? Rather than treating it as if it were one 12-lane superhighway in two parts, perhaps we should design them to perform distinct functions.

Zef Wagner is an analyst at Fregonese Associates and is pursuing a Master of Urban and Regional Planning degree at Portland State University, specializing in transportation planning. His views are his own and do not necessarily reflect the views of Fregonese or PSU.

22 responses to “Pros and Cons of Center- vs. Curb-Running Bus Rapid Transit”

  1. With this weekend’s closure of Powell at SE 21st, and seeing what it’s done to east west traffic throughout SE Portland, we should be careful of what we hope to see with a SE Powell line. That may not be the best route to add any mass transit to.

  2. I encourage anyone who that BRT is somehow less controversial than Light rail, look at whats going on down south in Eugene with their attempts to expand their EMX BRT system

    Thats not to say that Portland Metro shouldn’t do BRT in some areas, just don’t expect it to be any less controversial than light rail.

  3. Also, you do a great job illustrating why center lane would be a much preferable alternative to curbside. Thats a good test to see if a system will be true BRT, or something less.

  4. Dave,

    I don’t recall the first segments of EmX (the initial segment between downtown Eugene, the U of O, and downtown Springfield; and the Gateway extension) being that controversial–how come the W 11th BRT is different?

    Is it because West Eugene is more auto-centric and/or politically conservative than the rest of the area? How much of the opposition is due to concerns about the technical merits of BRT, as opposed to rants about socialism, crime, and other right-wing dog-whistles?

    Does the cancellation a while back of the West Eugene Parkway have anything to do with it? Any correlation between opposition to the BRT and support/opposition for the parkway?

    Is it due to concerns that the BRT might reduce auto capacity on W 11th (which is the current route OR126, a major state highway, and a frequent source of traffic jams due to the intermixing of regional and local traffic, something that the WEP was intended to remedy, its faults notwithstanding)?

    Have the initial segments of EmX noticeably underperformed, poisoning the well for future projects?

    Or was public outreach on the project simply poor and/or ham-handed–“this is what we are going to build, like it or not”?

  5. Scotty,

    I think its a combination of West Eugene being more conservative and auto-centric combined with the public outreach part of the deal. It seems the preferred alternative by those businesses is to simply do nothing, but at what point does congestion and service become so bad that you have to do something regardless (the polls say that Eugeneians slightly favor the route as a whole, but its awfully close).

    I don’t live in Eugene anymore, so I only know what I read in the Register-Guard, but my understanding is that the two existing EmX segments have been successful and improved ridership on those routes considerably.

  6. If you’re going to create a center lane style thoroughfare for transit, why not just install rail? Rail pays for itself over time and provides better service.

  7. I had heard the reason Interstate MAX could be done was because auto volumes were relatively low considering the available space, and the easy alternative of I5 provided access for longer trips.

    I’d love to see a similar conversion of 82nd avenue. Bus volumes are very high, and the parallel I-205 is only 10 blocks away. Unfortunately, volumes are pretty high on 82nd, and it would likely be very unpopular to drivers.

  8. Reggie,

    Depending on the what’s being installed, bus lanes/busways over existing street ROW are usually much less expensive than rail.

    * The roadbed is already designed to support the weight of bus. It certainly isn’t designed to support light rail. Streetcar-class vehicles (which present an axle load similar to bus) can generally work with roadbeds designed to support trucks and busses.

    * Utility relocation is also an issue for rail–utilities under the street (water mains, sewer/storm system, electric, gas, telco) generally shouldn’t go under tracks, as maintenance of the utility requires closing the rail line. (A bus can steer around if necessary). There are ways to make it possible to maintain utilities under tracks to make maintenance possible with the line remaining in operation, but that’s expensive to retrofit.

    The cost of a new rail line vs a new busway are similar; but for existing installations, bus can be much cheaper–in some cases, the cost of platforms and paint.

    The big problem is drivers and road authorities frequently object to the loss of road capacity.

  9. Scotty hits all the right points. I would add that people seem to think that acquiring right-of-way is the biggest cost of building rail, when it is actually a very small fraction. Even for Portland-Milwaukie, where lots of ROW had to be bought up, that was a pretty minor fraction of the total cost. In the case of center-running bus lanes, of course, the ROW is usually already in public hands–we just need to be willing to reallocate away from drivers.

    I also want to add to Scotty’s comments about the roadbed needing to be upgraded to handle rail. In the case of Barbur, I have been told that the viaducts would probably have to be rebuilt to handle light-rail. Now, it would be nice to rebuild them anyway since they are old and could use an upgrade, but that’s extremely expensive. BRT would not require all that extra work.

  10. Thanks, Bill. That really is the question–do we want these streets to be human places, or just car places? Powell is an important transportation facility, but using it purely to move cars is a pretty inefficient use of space and does make it lack a human scale. Dividing up the space into several sections (sidewalk, bike lane, car lane, bus lanes, car lane, bike lane, sidewalk) where possible would dramatically improve both the streetscape and the total person-moving capacity. Drivers will always complain about any change, but that shouldn’t stop us from making things better.

  11. As I’ve mentioned previously when the subject of BRT has arisen, Powell would be an ideal BRT “beta test” corridor (it makes sense politically since Gresham is already served by rail). And if they’re gonna do BRT, they might as well do it right and give it its own ROW, whether in the center or along the curb.

    Removing the US highway designation from Powell might help its transition into a multimodal local transit corridor; maybe 26 can join 99E to 224 to the future Sunrise Corridor (if they want to retain any designation for Powell, make it Business 26).

  12. Given the wide ROW to the east, I think a BRT line makes sense. Take advantage of it and establish center-running BRT that would share lanes with general traffic on the narrower stretches of Powell. If/when ridership increases, the gaps can be filled in with viaducts, tunnels, or land acquisition, and it can be converted to LRT.

  13. Interstate Avenue was no longer a state highway when the city made it available for LRT. And it was in bad shape and got a total rebuild as part of the Interstate project. The local match for it came from Airport LRT and Streetcar, so it was virtually all done with federal money.
    The CAC was very active in design (I was a member). Four stations were moved and/or redesigned at our insistance. The community demanded that the trackway be paved and Charlie Hales agreed. One of my regrets is that we agreed to pull the bike lane between Killingsworth and Rosa Parks stations; that was at PDC’s request so that retail parking could be retained at those station areas. Then, New Seasons then put is parking lot at the corner!
    Street design was another failure. PBOT, rather than placing the curb at the edge of the traffic lane and then cutting back where parking was possible or desired, put the curb parallel to the sidewalk. Hence there are many areas that are striped to indicate no parking, no bikes, no motor vehicles, no nothing, just dead spaces of asphalt. The Polish church opted to keep their trees and give up a couple of parking spots; a pity there was not more of that or that the default for extra street space was not to the parking strip rather than to asphalt.
    That said the Yellow Line MAX carries three times the riders of the old 5 bus and still does not get to Vancouver, WA. And Interstate Avenue looks a lot better than it did, and development continues slowly but surely. Just as we wanted.
    re cost of ROW, yes its cheaper if you can run an exclusive transit lane on an existing roadway, but even if MLR used McLouglin, it would still need the River bridge and several viaducts, hence its high cost, regardless of vehicle type. The same could be true for SW, especially as Barbur is mostly four lanes with only 5 or 6 between Naito and Beav-Hillsdale Hwy. My guess is taking two lanes for real BRT is a non-starter on Barbur.

  14. It seems like a vital piece of all center running configurations presented here would be the removal of auto lanes, and the remainder of one auto lane in each direction.

    This seems to be how the best cities in Europe do it: one lane each direction for autos, paired with dedicated transit lanes, dedicated cycle tracks, and sidewalks.

    But here, in the US, we seem to think one lane in each direction is inadequate on arterial streets. Is this true? Or do our “mobility” standards require that we oversupply auto capacity to ensure free-flow speeds?

    Multiple lanes of traffic seem like a deadly waste of space.

  15. I would like to see a cost comparison between a BRT running the length of the Powell/Division corridor vs. a LRT loop down Powell and continuing on the Green line.

    In the latter Buses could remain local from Gresham to I205 (and not have to worry as much about overcrowding). Both Division and Powell buses would have a frequent transfer to two MAX lines, the East Side Loop, and the Green line.

    In my experience longer bus routes have a tendency to deviate from schedule. Instead of relying on one bus in traffic, the system has more flexibility. Several bus lines would be effected by this, not just the 9, 4 and 14.

  16. A big question is how would rapid buses get across the UPRR- would they merger at SE 21st into the existing lanes that back up during rush hours? It seems that center lanes are preferable- curbside business access transit lanes do not achieve much for transit. Where the ROW is wide, between 52nd and 80th, there is not much delay. Also, the switch between Powell and Division on 82nd will add several minutes of travel time, due to the extra distance and two additional turns. Another option is to improve both to the present Division and Powell lines and not try to combine the two into a single line. Tri-Met could also go to all-door boarding on this and other frequent lines.

  17. “But here, in the US, we seem to think one lane in each direction is inadequate on arterial streets. Is this true? Or do our “mobility” standards require that we oversupply auto capacity to ensure free-flow speeds? ”

    I’m not so sure that it is a deliberate planning policy, as opposed to simply a vestige of times past in Europe. Most people living in European cities could barely afford a horse, let alone a comparatively, road hog carriage so they probably didn’t figure they needed wide streets. ‘Xcept in the fancier towns where they built wide boulevards with grand schemes in mind. When you say “best cities” how do we know you’re not stacking the deck with certain examples? :)

  18. Scotty,

    All you have to do to know the answer to your question is to look at the “MoneyPit” on the dragon bus caricature. It’s pure white-wing mouth-breather not-even-a-dog-whistle just-outright-pandering.

    I am hoping that the Iranians splatter Ras Tanura putting oil at $300/bbl and gas north of $10. That’s what it will take to wake Boobus Americanus up.

    Especially if it’s President OvenMitt after January 20th.

  19. The historic old city centers in Europe may be narrow, but in my experience as soon as you get farther out they have much wider right-of-ways than we have, or at least comparable to our suburbs. The difference is simply in how they allocate space. Sometimes you get the Paris-type boulevards, with through traffic in the middle and transit & bikes on the outside frontage lanes. Sometimes you get the Dutch-style transit in the center, then car lanes, then bike paths. In either case, it is better at accommodating modes and breaks up the streetscape with planted medians. We have many streets in the Portland area that could be retrofitted in this manner–it’s not just some European-specific thing.

  20. Some pros of curb side lanes:
    -curbside, if you are waiting for bus you might be able to deke into a shop and grab a coffee or a paper while you wait. Out in the centre, there ain’t much to do

    -in cold climes, you might be better sheltered by streetside buildings curbside, rather than exposed to wind in the centre

    -stepping off a bus curbside seems a lot safer than stepping off centre street. I realize there are barriers and busstops and so forth in the centre lane, but I found the experience a bit disorienting stepping of a Toronto streetcar in mid street.

    Mississauga is planning centre stops on busy bustling, fast-traffic Hurontario street. I’m not really looking forward to it. The northbound and southbound “stations” will be alternating (not opposed) leaving them exposed to wind and traffic noise on one side.

    -Dusseldorf has centre run transit. If you want to cross the street there can be two or three traffic lights to obey: the first to cross traffic lane to the LRT island. 2nd line to cross the LRT rails. 3rd lite from LRT island across the remaining traffic lanes. To a perpetual jaywalker it seemed a bit frustrating

    -odd how an ancient german city has wider boulevards which can handle LRT better than my hometown Toronto or Brampton. Bit baffling really.

Leave a Reply

Your email address will not be published. Required fields are marked *