Thinking about solutions for the Beaverton-Wilsonville corridor

A discussion of possible solutions to the Beaverton/Wilsonville corridor.
Back in February, we took a look at the future of commuter rail in the Portland area. One corridor which was examined in detail is the Beaverton-Wilsonville corridor, which has been identified by Metro as a “near term priority corridor” for high capacity transit. The corridor is presently served by the WES line, as well as by TriMet’s 76, 78, and 96 busses, and the 2X line of SMART. However, none of these lines provides adequate rapid transit service. Only the 76 and 78, both local bus lines, run seven days a week. The portion of the system where 76 and 78 overlap, between Beaverton and Tigard, has near-frequent service on weekdays, but the quality of the service offering drops off considerably south of Tigard. Clearly, the current service in the corridor isn’t where planners would like it to be.

But how to improve it?
The current thinking

The Regional Transportation Plan has called for WES service to be expanded to 15-minute all-day service, but given the steep operating cost of WES, and the need for the Portland and Western to move freight on the tracks, this may have difficulties as well.

How difficult?

The detailed analysis contained in the evaluation report (see pp 157-159 for the summary) included the following comments:

Given the recent opening of the WES commuter rail line, this corridor is unique to all other corridors considered in this analysis. Upgrading WES to 15/15 headway in 2035, consistent with the existing RTP financially constrained system plan, would require significant capital investment. There are significant physical challenges to upgrading WES to light rail or all-day, high frequency service. Freight rail will continue to operate in this corridor for the foreseeable future, requiring any significant update (to 15 minute or better all day service) to be constructed on a separate two-track line. Modeling of this corridor shows that travel demand is high and that an HCT investment on the corridor segments between Washington Square and Bonita are among the highest segments of any corridor modeled in this HCT System Plan. Other options for serving these segments should be explored in planning and design phases of the two Regional Priority corridors identified in this Plan – Corridor 11 (Portland to Sherwood via Barber) and Corridor 29 (Clackamas Town Center to Washington Square via RR ROW [Lake Oswego]).

Commuter rail is assumed to be the primary mode for this corridor. The highest demand
segments of the corridor could be developed as part of the regional light rail system using light rail as discussed in the Corridor 11 and Corridor 29 summaries.

The first obvious question: If you are going to building two additional dedicated tracks for transit service; tracks which will be separate from freight operations–why would you choose commuter rail over light rail? Even if you were to maintain long stop spacing, similar to what WES offers today, I have a hard time thinking of any technical reasons why operation of WES-style commuter trains over dedicated tracks would pose any advantages over light rail vehicles. The whole point of WES-style vehicles is they can run on existing active freight lines; but the downside is that they are subject to FRA regulations, which are demonstrably unfriendly to transit operations. And the Beaverton/Wilsonville corridor would benefit from additional stops along the way, particularly at places like Bridgeport and Washington Square, where the current tracks don’t go but a new light rail line could.

Light rail lines along this corridor could also through-run to Portland or Hillsboro along the existing Westside tracks, or switch to the Southwest Corridor, if the latter is built as LRT.

But the downside of doing this is the cost: Right now, adding two additional tracks in the corridor is pegged to cost about $1.9 billion–and given that this is a planning estimate, I suspect it’s on the low side. The corridor is nearly twice as long as Milwaukie-Portland, and while such a project wouldn’t include any bridges on the scale of the OMSI-SoWa crossing, one crossing of the Tualatin would be needed.

But the presence of existing rail service on the corridor may make another option–bus rapid transit of some sort–a (politically) viable alternative.

The bus/rail gap
Much has been written about the so-called bus/rail debate, in which the merits of busses vs trains in transit applications are widely discussed. As an abstract debate, it easily tends to devolve into an argument of platitudes, and strong pro-bus or pro-rail positions are frequently proxies for arguments about other values not having much to do with transit performance. That said, there are some applications where one technology may be better suited than the other.

For a cost-constrained system, especially where capital funds are limited, BRT has definite advantages. BRT systems generally cost less than equivalent rail systems, and at the lower end, can cost a lot less. The cost savings come from several areas:

  • “Difficult” parts of a corridor can often be dealt with by limited stretches of mixed-traffic running and bus lanes/busways can often be utilized by local bus services as well as by rapid services. Trains, on the other hand, require dedicated infrastructure wherever they go.
  • If the political will exists to convert general purpose auto lanes to exclusive bus use, significant increases in transit performance can be had at little cost; and the improved running times will help the operational bottom line, while making the service more attractive. However, if such conversions are not politically possible (which is often the case), then BRT will either require construction of a dedicated right of way (and thus be more expensive to build, with costs approaching rail), or require significant running in mixed traffic (and thus be slower and less reliable).
  • Roadbeds are cheaper to install then rail lines. A railbed capable of supporting LRT vehicles must be built to far stronger standards than one that needs to support busses (or streetcars, for that matter), and frequently requires relocation of utilities under the right-of-way. Train traffic can damage long runs of pipes underneath, and repairs to utilities located under rails may require taking the rails out of service. Busses do not pose a threat to utilities and can more easily detour should maintenance be required. On the downside, busses are notoriously hard on pavement.
  • It’s far easier to mix express and local services with busses–all are needed are passing lanes at platforms. Rail generally requires additional lines, or complex networks of pullouts and switches, to support an express service on the same corridor as a local.

BRT has several issues, however, which work against it.

  • Higher operating cost per passenger on high capacity lines. Trains can simply haul more people per driver than can busses; the individual vehicles are larger, and they can be entrained. This is only an advantage when you need the capacity. A MAX train costs about twice as much to operate per hour than a 40′ bus does; but can hold 4-5 times the passengers. A full MAX train is far more efficient from an operational point of view than the equivalent number of busses; but an 1/4 full MAX train is not.
  • Bus is frequently seen as less prestigious than rail by politicians–this consideration is perhaps more important than it ought to be, but it is real. After the North-South MAX line was defeated in the 1990s, BRT was examined for the Milwaukie corridor–only to have the city of Milwaukie insist that rail be put back on the table. (A discussion on the merits of BRT in the Milwaukie corridor is here, and many old Metro documents can be found here). The annals of transit development are filled with stories of activists and officials whining about how a neighboring community got “better” transit infrastructure. (When the Gold Line light rail opened in East LA last year, many locals complained bitterly that they didn’t get a subway like the west side got).
  • One of the most difficult objections, however, is the some-people-just-won’t-ride-a-bus argument. It’s a difficult topic, as it generally lies within the realm of psychology and culture (and can hit on some politically sensitive subjects such as race and class, that we won’t get into in this article); not within science or engineering. But there is some evidence that–regardless of the merits of this position–this gap is real; and that rail projects can expect higher ridership than equivalent bus lines. This is an important factor for an agency which sees improving ridership as a goal, whether for environmental reasons (getting more people out of cars) or for reasons of funding or prestige. Rail critics often argue that agencies shouldn’t worry about attracting finicky patrons, and should focus instead on those patrons who don’t mind the bus. For better or worse, though, TriMet cares about ridership, and the South Metro SDEIS claimed that light rail would attract 33% more riders than the top-grade BRT solution considered (with an exclusive right-of-way for busses). That said, the SDEIS projections for both modes far exceed the current FEIS ridership projections for Milwaukie MAX, so the figures might be taken with a grain of salt.
  • Rail also seems to be more attractive to developers. Whether or not this should be a consideration for rapid transit service is also a major question–a good argument can be made that if the developers aren’t contributing to a project’s funding, through a Local Improvement District, urban renewal, or provision of land (such as Bechtel’s involvement on the Red Line), etc.–that their concerns ought to be irrelevant. Like it or not, real estate interests have significant influence on policymakers, however.

TriMet has considered BRT for several projects in its history–the original MAX line was originally visioned as a busway, and the history of MLR is indicated above. However, it has never built any. Some allege an institutional bias against bus for rapid transit applications; some go further than that in their criticism.

But the Beaverton/Wilsonville corridor provides some unique opportunities.
The westside opportunity
Given these issues, BRT projects–particularly to the suburbs, where municipal governments compete for prestige and residents are more likely to be adverse to riding the bus (and developers leery of projects where proximity to bus service is a feature)–often come with several strikes against them. However, the WES corridor may be a promising one for the technology for several reasons.

  • Existing rail service in the corridor–WES. The presence of WES helps to address both the political prestige problem and the folks-won’t-ride-the-bus problem. Those who won’t ride the bus can ride the train (patrons who are bus-averse are more likely to only consider using transit for their daily commute, which is when WES runs); and the local governments involved already have a “prestigious” rail project running their jurisdictions, and thus far less room to complain about not getting another.
  • Likewise, a BRT line is more easily positioned as a complement to WES–a project widely viewed as a failure–allowing regional leaders to save face more than they would were a MAX line built instead.
  • Improved bus service would make WES more attractive for another reason–it would give those who work late (and might miss the last train) a more attractive fallback position.
  • BRT could continue to run on the 76 corridor, albeit with somewhat greater stop spacing, but still serving many important destinations that WES does not stop at.
  • With WES taking care of the peak loads on the corridor, the residual capacity needs might well be in the operational “sweet spot” of bus service rather than rail service.

In many ways, a BRT service in the corridor would be an excellent complement to the existing WES service, without the need to engage in a costly and difficult expansion thereof. And anything would likely be a competitive offering given the terrible traffic situation in the corridor–a while back, ODOT actually considered closing all the ramps on OR217 during rush hour so traffic could move more swiftly between US26 and I-5. Widening of 217 has been discussed–but the numerous closely-spaced interchanges make any improvements of the freeway an expensive proposition.

BRT wasn’t really considered as an option by the HCT plan, however. The evaluation process used in the HCT plan treated all proposed corridors as MAX lines (this was done to have apples-to-apples comparisons; several corridors are not recommended for light rail by the report, but may be suitable for other modes). The Beaverton/Wilsonville corridor was identified as suitable for rail; the possibility of a combination of commuter rail and BRT was not addressed.

What should it look like?

We’ll ignore the choice of bus or rail (and what flavor) for now, and look instead at service parameters, such as “how fast should it go”? and “where should it go?” (Another important question is “how often should it run”; however, I’ll ignore this question for now as I suspect any solution will be constrained by operating costs and not by line capacity).

The two questions, unfortunately, are not independent–a line that serves dense neighborhoods (particularly existing ones) is likely to be slower than a greenfield line, unless the community is willing to spend the money for things like grade separation. A line that is designed for speed will (by necessity) stop less often and be less-well integrated into neighborhoods–surface rail down city streets, even in medians, simply can’t run at 55MPH.

This is demonstrated by this table. While bus and rail are capable of similar speeds, all else being equal, high-speed transit (of any mode) requires that vehicles stop less often–and reliability requires that the number and duration of stops be predictable. The fastest services will only stop at platforms (class A); slower services may need to stop at a fixed number of crossings as well (class B); and services that run in mixed traffic may have to stop arbitrarily often, making them slow and unreliable.

How fast?

The first question, then, is “how fast”? For now, we’ll limit our discussion to the Beaverton/Tualatin segment, as that segment is entirely urbanized, and all within TriMet’s service territory. The existing service, WES, is fast–covering the ~9 mile rail distance from Beaverton to Tualatin in 17 minutes, and the full 14.7 mile route to Wilsonville in 27. Over the length of the run, it averages over 30MPH, which is very fast for transit. However, the average distance between stops is about 3.7 miles. On the other end of the service spectrum, the existing #76 bus (which only runs to Tualatin) offers travel times between Beaverton and Tualatin that range from 35 minutes during off-peak hours, to up to 45 minutes during rush hour. The distance of the route the bus takes is closer to 10.5 miles It takes about 32 minutes to drive the route taken by the bus, and 15 minutes (assuming no congestion) to use the freeway.

Doing faster than WES, while remaining useful, is not likely to happen. Marginal improvements could be had by vehicles not subject to FRA regulations, but the amount of savings are somewhat minimal, so we’ll consider 17 minutes a lower bound. Likewise, anything desiring to be “rapid transit” ought to be able to easily beat peak local bus service, and be comparable to off-peak service (empty busses that don’t have to stop can go faster, however)–I’ll arbitrarily set 37 minutes as an upper bound.

With 17 minutes offering a lower bound on the possible travel time, and ~35 minutes offering a reasonable offer bound

Where?

The route from Tualatin to Beaverton is somewhat diagonal, and the street network generally offers poor northwest-southeast connectivity in the region–with most major streets not headed to Portland being disjointed. The Tualatin River provides a barrier as well, as does the freight line, the freeway, and Fanno Creek. The main corridors where one could put a transit line are the freight line, I-5 and OR217, and Boones Ferry/Hall Boulevard. Grade-separated light rail (or BRT) would be most at home on the first two, though the 217 corridor in particular is narrow and might be problematic to expand.

The present 76 bus runs on the Hall/Boones Ferry corridor, other than a diversion to Greenburg Road between Tigard TC and Washington Square. This corridor hits all the major transit destinations, as well as serving numerous concentrations of apartments. Putting grade separation along Hall or Boones Ferry probably would be problematic, and disruptive to existing uses, but these roads could accommodate a surface line, particularly the stretch of Hall north/west of Washington Square.

The alternatives.

With all that in mind, here are five alternatives to consider, each offering different tradeoffs between cost and reliability, speed, and accessibility.

The 17-minute solution. Build a Class A (grade-separated; possibly with grade crossings guarded by gates) light rail line on a 2-track ROW from Tualatin TC, serving Durham/Bridgeport, Tigard TC, Washington Square, Denney/Allen, and Beaverton TC. (Actually, this line might follow the freeway across Canyon, and intersect with the existing MAX line near the big curve east of Beaverton TC and south of Sunset TC, with trains serving Sunset instead of Beaverton and continuing to Portland). This idea has two more stops than WES, but I’m making the assumption that the smaller and faster LRT vehicles can make up for it with better acceleration characteristics than the heavy DMUs used for WES.

The 22-minute solution. The same as the 17, but with more stops. To the list I would add a 72nd Avenue stop (near Carmen Drive), a Bonita Road stop, a Greenburg/Tiedeman stop, and either put back the Hall/Nimbus stop or add a Griffith Park/Farmington Road stop. Average stop distance would be right about a mile–similar to the Green Line south of Gateway.

The 30-minute solution. This would look a lot like the first two solutions south of Washington Square, with perhaps an additional stop or two (South Durham, just north of the river, is one possibility; a stop serving the Tigard Library might be another). But rather than following the WES line/freeway route into Beaverton, it would then run at grade in the median of Hall Boulevard, likely with a dedicated structure for crossing OR217 and Scholls Ferry Road (and the existing rail line), descending into the Hall Boulevard median near Nimbus–with stops at Nimbus, Greenway, Denney, Allen, and 5th. The tricky part is the routing through downtown Beaverton–Hall splits into a couplet north of Allen, and runs for half a mile through a residential neighborhood–there’s no place to put train tracks without blocking somebody’s driveway. There’s also limited space to widen Hall or Watson (and likely to be lots of resistance to converting existing traffic or parking lanes to rail, and a need for a grade-separated crossing of the freight line along Farmington.

The 35-minute solutionThe thirty-five minute solution is a good quality class B+ BRT line, largely running in the same route as the existing #76. But it has to be good-quality BRT (class B+), with an exclusive right-of-way and signal priority throughout. Exceptions might be made for downtown Beaverton, due to similar issues with the existing street grid. At Beaverton TC, a bus-only extension of Millikan Way could connect Hall/Watson to the existing BTC bus entrance, and special care would also be needed for bottlenecks around Washington Square and the crossing of 99W in Tigard. Hall and Durham south of Tigard TC might be tricky as well–these are three lane roads at present, and technically a state highway. (Given that Hall Boulevard no longer serves a significant regional purpose, a jurisdiction swap may be in order). Stop spacings could be a little closer than the light rail line, but not much.

The 40 minute solution. This is outside of the range specified above, but an EmX-style solution, with some exclusive lanes and some signal priority, could probably do the journey reliably in under 40 minutes or so–perhaps faster is some capital is spent to get around the worst bottlenecks. This may not seem like a big gain over the 45 minutes the local bus takes (and the greater stop spacing may annoy existing patrons who get a longer walk), but the reliability increase if it is done right is probably just as important as the speed. Stop spacings can be shorter, increasing the chances stops can be skipped. Reliability is one factor that so far has been glossed over, but it’s very important–if riders have to plan for the worst-case likely delay, that is often little better than a longer (but reliable) trip. And reliable services can have shorter turnaround times, permitting higher frequencies (another important issue not considered in this article).

For both BRT solutions, it is assumed that vehicles optimized for faster boarding (offboard/POP fare collection, multiple doors) are used.
Beyond the corridor

The above ignores extension of the line into Wilsonville, as well as service within Tualatin. The 76, after serving Tualatin Station, acts like a local circulator, serving the transit-friendly part of the city (roughly the square between the river, Boones Ferry, Mohawk, and SW 65th), with Meridian Park Hospital, an important transit destination, lying on the eastern edge of the square. But service to Wilsonville is also desirable. Right now, there is no transit service between Tualatin and Wilsonville on Sundays, and only the SMART 2X running on Saturdays. WES and the 96 provide service on weekdays. Seven-day service into Wilsonville would be beneficial, assuming a deal could be worked out with the city (it would permit the city to discontinue the 2X line). There are four reasonable corridors for such a line: The existing WES line (which passes on the edge of urbanized Tualatin, and may have growth potential), SW Boones Ferry (which passes through existing sprawl), I-5 (which has all the issues with freeway-running transit), and 65th, which partially lies outside the UGB (and is presently rural).

There’s also the question of some of the other corridors considered by the HCT plan, including the proposed Clackamas/Washington Square line and the Southwest Corridor. It may be possible that a partial solution is done involving one of these two (i.e. MAX from Beaverton to Washington Square/Tigard and then east to Lake Oswego and Milwaukie; or MAX from Portland to Tigard to Tualatin). Both of these projects are in the long term planning horizon, though.

Finally, there’s the possibility of extending service on the corridor north from Beaverton. The Sunset/Cornell Corridor is flagged as an important HCT corridor, and if a BRT solution is done, further upgrades north to Bethany or Cedar Hills might make sense. (I’m sure Al wouldn’t mind seeing the 67, for instance, getting a service upgrade, connecting BTC to PCC Rock Creek–even though this doesn’t appear to be on planning radar. And adding BRT infrastructure to SW Jenkins would have the additional benefit of speeding up the journeys of busses deadheading to Merlo Garage–right now, Jenkins is frequently a parking lot between the Nike campus and 158th; in both directions).

Thoughts?

Disclosure: I should note, in the interests of full disclosure, that my wife and I own a house near Hall Boulevard in Beaverton–some of the routings discussed do pass within several blocks.

21 Responses to Thinking about solutions for the Beaverton-Wilsonville corridor