Thursday, November 29, 2018

Why a Beltline Concept for the Bridges was Eliminated

With Councilor Andersen's proposal for a Work Session on the SRC adopted and in motion, in some quarters there is new talk of trying to build support for a beltline concept with bridges farther north and south.

River crossing at Lockhaven (1984, from below)
Such a plan has its own problems, and it's worth thinking a little about them. One set of problems arose through several planning processes and analyses, and accounts for why it did not appear in the draft Environmental Impact Statement among the vetted, formally recognized Alternatives. A different set of problems would not necessarily be recognized by any traffic study, and instead is more of a consideration about housing and land use, something that a Comprehensive Plan would address. Of course, they really are linked, not separate!

A Beltline doesn't Solve Traffic

You might have seen this "Salem Beltline" map recently. 

This concept may be from 1979 - via Facebook
It dates the concept back to a 1979 SATS Study (SATS being a predecessor of SKATS). The 1973 Comprehensive Plan does not show a beltline. It would be nice to find the earliest instance of it. It could date earlier than 1979. The relevant documents are mostly not archived publicly online, so here we will just sketch an outline. (And definitively establishing its origin isn't important just now.)

Moving a few years forward, in the "Year 2000" regional plan from 1984, there was indeed a beltline concept with bridges at Lockhaven Road in Keizer (detail at top) and southwest of Minto-Brown Park connecting with Kuebler Boulevard.

In 1984 we thought we'd need new bridges by 2000
(in red added; boxed comment in original)
But by the 2002 Willamette River Crossing Capacity Study, a Lockhaven/Chemawa alignment had been eliminated "due to the small amount of relief to the existing bridges." Traffic modeling, and the low congestion relief in that modeling, caused it to be discarded. They also identified Urban Growth Boundary issues as a barrier.

Lockhaven/Chemawa concepts eliminated in 2002
An analysis of a bridge on a Kuebler alignment was a little muddled, but they retained it:
A bridge at Kuebler would not effectively address the need of reducing traffic on the existing bridges. A Kuebler bridge would, however, serve as a bypass around downtown Salem, but would not be expected to significantly reduce the through traffic on the downtown arterial system. While large trucks may continue to be only a small percentage of the overall traffic flow, there will someday be a point when it will be desirable to remove these trucks and other through traffic from the downtown Salem street grid....[We recommend] retain[ing] the Kuebler Corridor for further consideration because of its ability to connect Highway 22 with Kuebler Boulevard and I-5, thereby providing an alternative long-term route around the downtown Salem core area.
The general beltline concept was also retained and the study deferred "the beltway concept for further consideration and study at some future date."

It is curious that while they retained "the beltway concept" in general, they also kept finding difficulties with key parts of it, and nothing in the analysis pointed to a beltline that actually would work. It remained a vague, theoretical hope mostly unsupported by each current study's modeling and analytics. It might be right to call it "a fantasy."

Still, they came back to it again.

In 2006, Nos. 1, 2, & 13 scored low on congestion relief
In the 2006 Study Area Refinement memo they revisited some of the concepts, but prospective bridges at Lockhaven, Chemawa, and Keubler (numbers 1, 2, & 13 on the map) scored low for alleviating "vehicle hours of delay."
The most obvious pattern evident from the v/c result is that the furthest south (Kuebler) and northern two (Lockhaven and Chemawa) corridors do the poorest job improving river crossing demand in the peak westbound direction.
If you accept the traffic modeling (and there are reasons not to! but for the moment in the argument here, we will accept the conclusions), a beltline concept with bridges at or near Lockhaven and Kuebler doesn't much reduce congestion.

That's why we haven't been talking about a beltline during the SRC process. It was not one of the Alternatives evaluated for the draft Environmental Impact Statement.

Induced Development on the Edges

Eugene has a beltline (click to enlarge)
There is another reason to be skeptical of a beltline concept, and you can see it easily in Eugene and in South Salem near Keubler. Even if, as would be with Cordon Road, you make the beltline coincide with the Urban Growth Boundary, it will be natural for a beltline to attract development on both sides. Why would you build an expensive system to serve one side of a road only?

A complete beltline will induce more sprawly, car-dependent development on the "other side" of it.

We see this clearly on Kuebler Boulevard also.

Making a choice for a beltline is to induce more traffic and more low-density development on the city edges. That's why a beltline concept even in general should not be attractive.

Part of our hydraulic autoism is that we think of an existing pool or stream of traffic that we have to relieve. Under this model we build a new road, and there is now free- or freer- flowing traffic. We have relieved the pressure. But mostly we exclude the new travel and new trips induced by the new capacity - the "new" pressure.

But as we know from the foundational paper, "The Fundamental Law of Road Congestion," "increased provision of...major roads is unlikely to relieve congestion." New trips and new development will fill the road and adjacent areas.

(It's still a free download!)

Drawing Lines on a Map

This floodplain map from 1973 is very clear
Still, when you are just "drawing lines on a map" and believe "solving traffic" is the most important thing, it's easy to see why a bridge at Chemawa or a beltline concept is attractive. The river is narrower and it's easy to extend a line at Chemewa.

But if you look at the floodplain, a span at Chemawa would still be wider than you might think. It's even wider just a little north of there, and Willow Lake and the wastewater plant also make a Lockhaven extension more difficult. (At Kuebler is another matter, and even more difficult; there the river's plain is wide.)

Over and over it's clear why our bridges are all clustered downtown. The geography of the river's waist is destiny! Those 19th century engineers knew a thing or two!
First bridge of 1886, probably from Minto looking north
(Salem Library Historic Photos,
but mislabeled as second bridge of 1891)

Do we let traffic drive land-use? Or ask land-use to drive traffic?

In the end it is likely that we need to focus more on land-use policy and planning than on traffic policy and planning.

Or, less abstractly, we can make congestion the primary problem to solve, or we can make housing the primary problem to solve.

And the essence of the matter would seem to be: Why aren't we rushing to make housing the center of our analysis? (So maybe "Our Salem" and not the SRC should drive the process and planning.)

Once we decisively put housing at the center of our analysis, we will look at traffic and congestion in a much different light. Broadly speaking, we can continue to maintain our exclusionary apartment ban and force new housing and development into greenfields at the edge of the city, which requires one approach to cars and congestion; or we can end the ban and focus new development on areas already well served by walking, biking, and transit, which entails a different approach to cars and congestion.

We need to choose one, and not try to satisfy both at the same time:
  • A traffic policy that induces development on the city edges
  • A land use policy that induces development nearer the city center
For these reasons, a beltline with its associated bridges is not an obvious solution.


Walker said...

There is no “solution” — traffic headaches caused by thoughtless design and unforeseen effects are a predicament that can be somewhat addressed (or made worse) but they are not finite problems capable of being “solved.” Nearly all attempts to “solve” a predicament end up addressing only symptoms, and usually magnifying the symptoms that cause distress.

As Chuck Marohn noted, most people demanding “solutions” are only willing to consider things that translate to “How can everyone else change what they’re doing so I can keep doing what I’m doing unchanged?”

The cry “Build me a bridge” is a good example — the delusion that Salemites should pour hundreds of millions (almost $1 billion before cost overruns are included) into a gargantuan bridge so that auto sprawl can resume deeper and deeper into Polk County is, at heart, the demand of builders and land developers that 1970s suburban mass production tract development be allowed to continue so that they can keep doing what they’re doing, at everyone else’s expense.

Anonymous said...

To add to your evaluation of the beltline concept, a major challenge is the topography for a Kuebler corridor. The corridor (from Hwy22 at Eola to Kuebler at Croissan Creek) is over 4 miles long (as the crow flies) and after spanning the flood plain it goes from an elevation of 200 ft (at River Road S) to 700 ft (top of the south Salem hills) before connecting to Kuebler Blvd near Croissan Creek Rd at the city limits boundary (elevation 440). That length of infrastructure and the elevation changes are particularly challenging.

Regarding housing, Oregon's urban growth boundaries have help to concentrate population, and that's not just an opinion. Look at this webpage with the list of 497 urbanized areas in the U.S. It may come as a surprise that Oregon's urban area have higher population densities (Portland, Salem, Eugene, and Corvallis are in the top 100 of 497), with Salem ranked 59th.

List of United States Urban Areas (2010)

I downloaded and sorted the list by population density (pop/square mile) with the results as follows for Oregon:

Portland - 3528 persons/square mile (rank: 34 of 497)
Salem - 3117 persons/square mile (rank: 59 of 497)
Corvallis - 2957 persons/square mile (rank: 70 of 497)
Eugene - 2852 persons/square mile (rank: 84 of 497)
Medford - 2381 persons/square mile (rank: 131 of 497)

(The Census's urbanized area for Salem includes the cities of Salem, Keizer, Turner and a few locations just outside the urban growth boundary
Map of Salem urbanized area (2010 Census)

Other urban areas that have a similar population density to Salem are Seattle (3028 persons/sq mile) and Baltimore (3073 persons/sq mile).