Archive | Bus Rapid Transit

I’m a Proud CTU

Via @LitmanVPI:

That would be a Cycle-Transit User, someone who combines cycling and transit to accomplish a trip. A new study out of the Mineta Transportation Institute (PDF) looks at this behavior in Philadelphia and San Francisco.

A few top line conclusions:

  • Cycling and transit act as access for each other, it’s not a one-way relationship
  • Cycling extends transit catchment distances to several miles, although not always in obvious ways. Travelers may use cycling to avoid a transfer, reach an interim destination not on transit, or other creative ways.
  • While the largest use case is taking a bike onto the transit vehicle, there are lots of use cases that depend on locking up the bike, and agencies can facilitate this combined mode by offering plentiful secure bike parking at key transit locations.

That last point is one we emphasized in the Portland Bicycle Plan for 2030 (I chaired the committee that worked on the bicycle/transit integration chapter).

Initial Thoughts on the ITDP Study

The transportation corner of the internet has been abuzz for the last week over the new study from the Institute for Transportation and Development Policy.

The report compares the amount of development catalyzed by a variety of transit projects, including Light Rail, Bus Rapid Transit and Streetcar – and looks at the ratio between capital investment in the transit system and the amount of development. It reaches the somewhat surprising conclusion that the best return on investment was from the Cleveland HealthLine BRT system.

HealthLine is compared against Portland’s own Blue Line MAX. While the Blue Line has a higher total development amount, the capital investment in transit is also much higher.

Portland Streetcar fares well with the third-highest return on investment.

The report suggests several conclusions about what makes for successful development along a transit line:

  • Run the line through a strong or emerging real estate market area
  • Make sure the line runs through the downtown district
  • Ensure that there are complementary government policies and programs that support development (zoning, comprehensive planning, investment tools, etc.)

These all make sense to me.

The startling conclusion the report reaches however is that the type and quality of transit matter very little:

“Our analysis found no correlation between the type of transit investment and the level of TOD investment. LRTs, BRTs, and streetcars all led to similar TOD investment outcomes under similar conditions.”

“The quality of the transit system investment matters, but only marginally. The very strong TOD impacts in both the silver-standard Cleveland HealthLine BRT and the silver-standard Blue Line LRT (Emerging land markets with Strong government TOD support) outperformed the TOD impacts of the below-basic BRT systems in Strong land markets with Strong government support (Kansas City Main Street MAX, Seattle SLU Streetcar, Portland Streetcar). This is the only clear indication that a higher-quality transit investment helped leverage more TOD impacts.”

The report has a methodology for evaluating the transit in term of “BRT Standard” equivalents, based on an elaborate scoring system, where Curatiba is Gold. The Cleveland HealthLine and our Blue Line are both rated Silver (there are no Gold systems in North America) while Eugene’s EmX is rated Bronze. Portland Streetcar is considered “below basic” (I’m hurt).

If you’re wondering about the choice of “BRT Standard” units as the scoring mechanism, keep in mind that WikiPedia indicates that ITDP “focuses primarily on developing bus rapid transit (BRT) systems.”

The reactions from different quarters have been interesting to watch:

So what do I think?

Fundamentally I agree with the idea that a combination of factors need to come together to stimulate development. But I can’t accept that the nature of transit doesn’t have significant impact.

Unfortunately, the report’s methodology (at least as reported) is not rigorous. An academic analysis of the same issues would include things like correlation coefficients, etc. We don’t have anything like that kind of quantitative analysis  here. And I suspect others are going to fact check a lot of details in the report.

I think the report suffers from the same “transit is about long trips” bias that imbues much of the criticism of Streetcar. Many of the reasons Streetcar scored “below basic” are related to not making an effort to support long, fast trips (but at the same time, Streetcar scores very favorably when compared on a riders/mile metric). For those of us working to create sustainable urban environments that rely more on short trips, Streetcar is a smashing success. I do agree with the report’s suggestion that Streetcar couldn’t work if there were not complimentary transit available for trips to other areas.

Fundamentally, the “type/qualify of transit doesn’t matter much” conclusion can’t be correct. My own hypothesis would be that what DOES matter about transit is the commitment of infrastructure in fixed locations that developers know won’t move anytime soon (permanence). Rails definitely serve that function, but it may well be that investment in BRT stations (the report describes “iconic” stations on the Cleveland HealthLine) and dedicated right-of-way is effective as well. Certainly here in Portland a number of folks have posed the question of whether catenary wires for electric trolley buses would have a similar effect. I hope at some point we see a rigorous study that illuminates this debate.


Portland BRT: How open should it be?

Bus Rapid Transit has been getting a lot of attention in the Portland metro area recently. North of the Columbia, C-TRAN has been planning its Fourth Plain BRT project. Here in Oregon, the Powell/Division Transit Project, widely expected to be BRT of some sort, is getting ready to start. (You might be interested in this presentation from TriMet from last year). Higher-end BRT is a strong contender for the Southwest Corridor, and longer term, BRT has been actively discussed for corridors such as Tualatin to Oregon City/Clackamas via I-205, and TV Highway between Hillsboro and Beaverton. And the transit mall, essentially, is a dedicated transitway (though the high stop density and large number of signalized crossings keep it from being rapid), and when the Portland-Milwaukie Light Rail project (and the new bridge) opens, there will essentially be a dedicated multimodal transit corridor stretching from OMSI to Union Station.

North of here, King County Metro operates the RapidRide service, and Community Transit operates the Swift BRT service. South of here, the Lane Transit District operates the EmX service in Eugene and Springfield. And in Salt Lake City, the Utah Transit Authority operates a BRT line called…. MAX. (Their light rail is called TRAX, for those who are interested). Yet so far, no BRT system is operating in the Portland metro area–but it appears that will change.

An important decision when considering BRT–especially when considering an expansive system, rather than a single line or two–how open should it be?
Open vs closed BRT

In rough terms, an “open” BRT system is one where service is decoupled from physical infrastructure, and a “closed” one is where there is a high degree of separation between BRT service(s) and POBS (“plain old bus service”). The high-end Brisbane (Australia) BRT system is a canonical example of an open BRT system. Grade-separated busways radiate out from the central business district, with many overlapping bus lines using the busways as a high-speed, high-frequency trunk, and then leave the busways and run on local streets to serve different neighborhoods. Ottawa’s BRT system is also an open BRT. A closed BRT system, on the other hand, more resembles a light-rail line; as dedicated busses (often with distinct branding and livery) travel on dedicated infrastructure over specific corridors, and are distinct from local bus service. EmX is a closed BRT system, as is the LA Orange Line.

Of course, “open” and “closed” is not a binary choice. Busses, fundamentally, are all rubber-tired vehicles designed to run on roads, so there is plenty of opportunity to mix and match. Swift is one example of a “hybrid” system–as several local Community Transit routes overlap with Swift, and are capable of using some of its infrastructure. In general, for a hybrid system, there are two questions to ask:

  • To what extent can “ordinary” busses take advantage of BRT infrastructure?
  • To what extent can “rapid transit” busses venture off the busway and serve neighborhoods via local streets?

Among the attributes to consider are:

  • Right-of-way needs of the vehicles: Are oversized busses used that may have difficulty running on many local streets? Is special guidance equipment, signalling, etc. needed to safely navigate a busway?
  • Platforms and platform compatibility: Are stations high-platform, low-platform, or curb-level? Is precision docking needed, to minimze the gap between platform and bus (and permit rampless boarding of wheelchairs and such)? Are on-board ramps needed? Are opposite-side doors (doors on the left in the US) needed for center platforms? Are separate platforms/stops provided for different types of service?
  • Fare collection/payment: How are fares collected–by the driver, by onboard ticket machines/validators, by ticket machines/validators at the platform, or by fare gates on the platform? Are fares enforced by physical security (gates), checked by drivers or conductors, or randomly checked by fare inspectors?
  • How easy is it for faster services to pass slower services, particularly for limited-stop services to pass local services stopped at stations not served by the faster service?

Hybrid system use cases

There are several specific use cases that are worth considering:

  • Combined rapid and local service: A common characteristic of BRT is wide stop spacing–there’s little point in speeding up bus infrastructure if the bus is to stop every 150m. 400m (about a quarter of a mile) is often specified as a minimum stop spacing, and many BRT systems average a mile or more between stops. While this does not pose a problem for many able-bodied riders–many of whom will happily trade off walking distance for faster, frequent, and more reliable service–it can negatively impact those with mobility impairments, for whom a longer walk is a burden. A common solution is to mix local and BRT service in the same corridor. LA does this with its local (orange) and Metro Rapid (red) busses; this is not difficult to do with mixed-traffic BRT (such as Metro Rapid) or BAT-lane BRT (like Swift). For higher-end BRT, however, mixing local service in can be more problematic. Several local LTD bus routes run alongside EMX on Franklin Boulevard in Eugene–but they serve separate (curbside) stops (and stay in the mixed-traffic mainline) whereas EMX serves BRT-exclusive stops in the median. As many of the EmX stops require left-side boarding, use of LTD’s regular busses on the EmX line simply won’t work.
    I can see this scenario being played out with the proposed Powell/Division line, with the rapid line having infrequent stations and queue jump lanes, and the 9 and 4 being maintained as locals (with lower frequencies) that can also take advantage of the queue jumps, but stop more frequently and may not stop at dedicated BRT stations, depending on if there are platform compatibility issues. Actually, any BRT trunk line is a candidate for a nearby or overlapping local.
  • Branching closed BRT: BRT doesn’t necessarily have to be open to take advantage of trunk/branch topology (one of the things it does that rail cannot do easily)–it just needs to be possible for BRT-specific rolling stock to travel off the busway (possibly with lesser performance). One important attribute to support this is onboard payment, if the bus will be servicing stops that don’t have ticket machines. On-board ticket machines (like Streetcar has) are a good way to do this (though machines at platforms on the BRT trunk should still be provided). Mobile ticketing apps, like TriMet is presently beta-testing, will work. A second important attribute is adequate fare inspection–you don’t want to make everyone file past the driver on a BRT service. (And mobile ticketing will probably require this as well). Many European transit agencies have use proof-of-payment on all their services, there are good reasons TriMet should head in this direction too.
    Branching will likely be important for the Southwest Corridor–simply because the existing bus service has a heavily branched topology, with numerous lines heading south on Barbur/I-5/Corbett and scattering to various destinations in west and southwest portion. Branches to Raleigh Hills/Beaverton, Garden Home/Washington Square/Murrayhill, Tigard/King City, are certainly obvious candidates. In addition, if BRT infrastructure is built between Tigard and Tualatin–might the 76 be converted to BRT as well, with the Tigard/Beaverton section done as an unimproved branch? Upgrading the northern end of the 76/78 to BRT would itself be a worthwhile project, though outside the scope of the SW Corridor.
  • Partial BRT: Another advantage of BRT over rail is that it can far more easily be staged–were a line to Tualatin built in two phases (phase 1 to Tigard, phase 2 to Tualatin), it is entirely possible for the service to continue all the way to Tualatin on surface streets in the interim, and upgrade it to higher-quality infrastructure later. And likewise for an extension to Sherwood. Or it may be the case that funding is only available for a partial line, and a street-running extension becomes a long-term solution. Or possibly “class C+” BRT at the extremities, and higher quality service closer in to town.
  • Express service on BRT. Every time TriMet opens a light-rail line, it tends to cancel redundant bus service–and a particular target of that axe are express bus routes. There are good reasons for TriMet to do this–rapid transit can generally provide similar levels of service for suburban commuters, and express busses are expensive, inefficient, and in some cases, inequitible–but many riders of express bus like the service, and don’t consider LRT to be an adequate replacement, for various reasons (some better than others). Some fraction of these former express riders can usually be counted on to abandon transit and switch to driving (and to complain about it in the press). With BRT, a viable option is to simply run the express lines on the BRT infrastructure, skipping many of the stations along the way. If this is to be done, passing facilities need to be provided, so express busses can pass locals or ordinary BRT servoce. If ordinary busses (or even coach-configured busses, like C-TRAN’s express lines) are used, the stops that they serve need to be compatible (curb-level platforms on the right) even if the ones they skip are not (raised platforms on the left).
    TriMet is planning to upgrade the 94 to all-day service, as part of its 2014 budget; providing an express alternative to the 12 between Tigard and Portland. The 96 and 92 also provide express service in the corridor catchment area (these latter two lines provide service for which there is no local equivalent, an unfortunate circumstance in my opinion). Whether this would be continued were the SW Corridor to be built (would we get three levels of service in the Barbur corridor?) is to be decided, but it is a distinct possibility.

Southwest Corridor Questionnaire

This past Thursday, May 23rd, a community planning forum was held to present and discuss the current status of the SW Corridor, and solicit more community input. For those who didn’t attend, the materials have been posted on the project’s home page, and there is an online questionnaire for those who wish to add their two cents on the project.

Of some interest was this set of slides showing trade-offs of different routings, and this slide, giving preliminary cost and ridership estimates for different variations of the corridor. The relevant details are repeated after the jump:
The numbers

Capital costs:

  • OHSU tunnel: $3.1B
  • LRT to Tigard: $1.7B
  • LRT to Tigard and Tualatin: $2.6B
  • “Best performing BRT”: 50%-80% of LRT ($850M-$1.36B to Tigard, $1.3-$2.1B to Tualatin)
  • “Mid performing BRT”: 40%-50% of LRT ($680M-$850M to Tigard, $1.0B-1.3B to Tualatin)
  • “Lower-performing BRT”: < 40% of LRT


  • LRT to Tigard: 22.5k
  • BRT to Tigard: 20.1k
  • BRT to Tualatin: 26.9k
  • BRT to Sherwood (via Tigard and Tualatin): 28.9k

Operating cost/year:

  • LRT to Tigard: $4.9M
  • BRT to Tigard: $6.3M
  • BRT to Tualatin: $7.5M
  • BRT to Sherwood: $10.1M

A few caveats on the numbers:

  • It is unclear whether the $3.1B for a tunnel is instead of a surface route, or in addition to it.
  • Ridership and operating cost for other scenarios not listed, has not been modelled.
  • I believe that ridership and op cost figures are the costs and ridership of the new service, and don’t take into affect changes to other services. (In other words, the ridership figures don’t appear to be all new riders, and some operational savings may be had from reducing redundant local service).
  • Route to Sherwood is via Tualatin; a King City routing appears to be out of the picture at this point.
  • A line to Tigard would terminate at Tigard TC; it is unclear where a Tualatin extension would end (the two logical places would be at Tualatin Station and at Meridian Park Hospital).


The immediate thought is that the capital costs, particularly for light rail or high-end BRT, are high. The distance between PSU and Tigard TC, with deviations to the Tigard Triangle and PCC-Sylvania, is about 10 miles, so the cost-per-mile for a surface LRT solution is actually a bit lower than PMLR. The West Hills cannot be an easy place to put a transit line–on the other hand, there’s no major bridge needed for this project (an extension to Tualatin may need a new crossing of the Tualatin River, but that’s far less of a big deal than a crossing of the Willamette. The distance of a route to Tualatin is nearly 15 miles. Assuming a higher-end BRT can be done for 2/3 the price, one could reach Tualatin for the same budget as LRT to Tigard, and still provide a high quality transit service.

Of course, these are all preliminary numbers; not a budget of the scope and accuracy that one would find in a DEIS or other advanced planning document. Unfortunately, preliminary numbers tend to be on the low side…

BRT would provide a few other opportunities–it supports branching service much better, and it’s easier to phase: Assuming the vehicles used aren’t too exotic, busses on a BRT going to Tigard can continue on surface streets to Tualatin if there isn’t enough funding to go all the way to Tualatin in one go–and likewise to Sherwood. By the same token, it may be possible for the 76 to use a new busway or bus lane between Tigard and Tualatin. (And while this isn’t in the works for the SW Corridor, a BRT between Beaverton, Washington Square, and Tigard is a tempting idea…)

The big question is–where and when will money come from to build anything on the scale of what is discussed above? And the second question–which I’ve asked several times before, but not really seen a satisfactory answer to–why have capital costs for this sort of thing skyrocketed, particularly in comparison to earlier MAX lines?

The floor is open.