Willamette bridge: examining the options, and estimating the estimates (UPDATED)

The Willamette River Bridge Advisory Committee meets Thursday to look at the three bridge finalists: a cable-stay, a refined cable-stay, and the Miguel Rosales hybrid concept. Actually, you could call it four finalists: the cable-stay has a 2-pier option and a 4-pier option.

Let's break things down from there.

First, height: the two traditional cable-stay options (pictured below) would produce bridges of either 277 feet or 254 feet (for the 4 pier or the 2-pier option). The refined cable-stay option and the hybrid (pictured above) would each, on the other hand, be 180 feet tall. And we want this bridge to be not so tall.

To be fair to TriMet, which seems interested in using a version of the traditional cable-stay bridge for its greater track record and familiarity, has found an option in the refined cable-stay that achieves the same lower height as the hybrid that gained favor with the public, city leaders and design community.

In terms of aesthetics beside height, the refined cable-stay would also have a different, arguably more contemporary look from the other two cable-stay options, with open-ended V-shaped columns. However, the V shape is common in bridge design today, evidenced by projects around the world with that look. The V look might be attractive to many, but it's not unique.

The refined cable stayed bridge is mainly a bad copy of the hybrid bridge with the added problem of the inclined towers, which are very inefficient and counterintuitive.  Why would you incline the tower toward the walkways and encroach into the space of pedestrians and bicyclists which is already at a premium?

Some readers have taken issue with my frequent use of the phrase "cookie-cutter" to describe the cable-stay options. After all, if a cable-stay type can be built for this Willamette crossing without any more unwanted height than the Rosales hybrid, why be so against it? And indeed it's true: you can make an excellent bridge out of a common type.

But the hybrid is actually different in that it was designed specifically for this site, for this crossing. It takes the forms of the other bridges into account, such as the geometry of the adjacent Ross Island and Marquam. It's rare in this era to get a bridge uniquely tailored for the site, but the hybrid is it.

One can't easily discern the differences between the hybrid and the cable stay in small JPEG photos like those posted here. It's a matter of nuance. But ask most people in the design community and they'll tell you the hybrid is a better fit. The Portland Design Commission wrote a letter of support for Rosales' hybrid concept.

Now, let's look at cost. TriMet will soon announce the following estimates for the three main bridge types:

cable-stay (4-pier): $82.66 million

refined cable-stay: $89.23 million

hybrid: $112.78 million

The transit agency's numbers for the first two, the cable-stay and refined cable-stay, are probably sound. As noted by TriMet in previous releases, they used the National Constructors Group (NCG) for an independent cost estimate to build the wave frame and cable-stayed bridge type. The NCG  has been used by government agencies throughout the U.S. to do independent analysis on the cost to build major infrastructure, including bridges. What's more, they are familiar with cable-stay bridges. But perhaps I shouldn't say "they," becuase the NCG estimates are really only done by one self-employed guy in San Francisco, who is also reportedly a good friend of the Willamette bridge's main engineer from HNTB, a retired engineer who works out of his house and doesn't even have a website.

Conflict of interest? I'm just asking.

The hybrid funding estimate is also (or maybe I should say "therefore") very debatable. According to a source I spoke with on the condition of anonymity, TriMet is estimating 25 percent greater cost for the hybrid when in reality it's probably something more like 10 percent. And as previously mentioned, Miguel Rosales and his team told Portland Spaces they have not been contacted for any detailed info on the hybrid.

When the Rosales and Partners/Schlaich Bergermann und Partner team estimated the costs of their own hybrid proposal, they came up with $93 million. That's a BIG difference from the $112 or even $119 million estimates given by people outside the design group, all of whom have not contacted Rosales or Schlaich Bergermann for cost info.

Let's put it this way: In this stage of the design process, no matter who is doing the cost estimating, you can create a cost estimate for this bridge that is somewhere between 80 percent and 120 percent of what the actual cost will be. It's all just guesswork, but within that lies the opportunity to fudge the "estimate" in whichever direction fits the choice you already want to make.

To give this process real integrity, the city should hire a completely independent cost estimator.

What's more, the source points out that the refined cable-stay's V-shaped form could easily wind up taller than the estimated 180 feet. The incline of the spires reduces the stiffness of the cables and therefore their ability to provide structural support. That means more  steel cable is required to achieve the same stiffness as there would be with vertical towers.

I'd be happy to offer TriMet a chance to refute this. But instead of refuting the info I'm passing on, they might want to start by checking their numbers on the hybrid with the hybrid's designers.

Ultimately, though, I worry it's just not in the cards for Portland to get a uniquely tailored bridge as long as TriMet is in charge of the process. These are not bad people. TriMet does a lot of good things in Portland. More than practically any of our buildings, the city's design emblem is its mass transit system, particularly MAX and the streetcar. But TriMet is a transit agency. They are not seekers of great bridges for the city often called "Bridgetown". Their task is to build a bridge for MAX across the Willamette with as little risk and surprise as possible.

Perhaps t do something great, we need leadership at the City level and from a constituency of citizens demanding that design excellence be a priority.

Comments

[Nate G.]

Thanks for the super informative post, Brian.

Just from looking at those cost numbers and the way the new renderings have been released by Tri Met, it's hard not to feel like the fix is more or less in for the "refined cable-stayed" option.

Do you know, off the top of your head, of any other specific examples where the V-shaped cable-stayed design has been built elsewhere in the world? I'd love to see one from a different angle than the one offered up by Tri Met (which the cynic in me thinks may have been chosen specifically to minimize as much as possible the bridge's differences from Rosales' design).

Until more renderings come out, I think I'd have to say that I'd be thrilled with Rosales' hybrid, seriously disappointed with the "unrefined" cable-stayed, and cautiously optimistic about the "refined cable-stayed."

Thanks for all your reporting on this story.

[Scott Mizée]

I'm not sure if you saw this, but I just came across this interview with Donald McDonald posted on Tuesday: http://bit.ly/dHLZE

Q&A: Donald MacDonald, Architect
Donald MacDonald has designed more than a dozen bridges since establishing his architectural firm in 1966. Today, he has bridges scheduled for construction in Dubai and Portland, OR, and he's in the midst of building the world's largest self-anchored suspension bridge:

[Aneeda]

I don't mind the taller bridges...maybe it is the renderings, but they seem much more elegant than the hybrid or the refined cable stay. None of these bridges are that unique and the taller bridges have a much more elegant connection to the river in my opinion. The shorter bridges don't really seem to resolve themselves in the air very well.

[Jonathan]

For the sake of comparison here are the heights of three existing bridges:

Hawthorne Bridge: 165 ft
Fremont Bridge: 381 ft
Marquam Bridge: approx. 160 ft (top of upper deck) - not totally sure about this number

Has anyone looked at the comparative scales of these bridge...like built a model of all the bridges along the river to understand what implications the scale might have?

[Don Dulyea]

Seeing a model with the heights of the other bridges might be helpful...especially to see how much contrast would be had with the Marquam Bridge.

If the two were about the same height, would the view of downtown be obscured by that big concrete road in the middle of the air? If so, this would of course be an issue with any column height, but to see both from afar, it might be better for the new bridge to claim some prominence over the Marquam.

Although, it has to be said, the view from the upper deck driving north yesterday evening at dusk almost sent my car into the guardrail. You definitely have to appreciate a clear, unobstructed view.

[pylon]

Thanks for laying things out, Brian. Thou rocketh.

I too would like to see other real world examples of the V-shaped cable-stayed refined. BTW, there's nothing extraordinarily unique about the two bridges that would bookend the new bridge (i.e., Marquam, Ross Island). Perhaps a bit of contrasting daring-do would add visual spice to the river. Cities like Barcelona, Paris, and Amsterdam have been successfully contrasting the old with the new (i.e., avoiding trying to merely blend in) for quite awhile now- letting the fresh new designs signal a break with the past.

And lastly, if I were a regular user of the bridge, I'd like to have as little cabling around me as possible while crossing in order to create a more open "floating over the river" feeling, as well as provide a more unobstructed view. The hybrid seems to offer the most cabling.

[Bob R.]

It's my understanding from other renderings I've seen that the cabling would be _between_ the ped/bike lanes and the primary transit lanes, so you'd have an unobstructed view in at least one direction when walking/biking.

[billb]

can we paint this any other color than white , subtle it is not.

[ac]

the four pictured don't represent much of a range of aesthetic options...they're very similar...so much so that i'd be happy picking by cost at this point

[Brian Libby]

ac,

I wouldn't be so quick to judge nuances of design from little JPEG images. I wanted to post these images to give people a taste of the bridges, but trust me - they require a closer look.

[Steve]

I would like the towers to be less angular and instead curved like an elongated teardrop. Perhaps incorporating a structural “scissor tail” at the top of the towers extending above where curves intersect.

[Scott]

I like that idea, Steve...and it would be nice to have some more "organic" architecture in PDX...

[ws]

Cable Stayed-refined (top picture)...don't like it. That 'V' thing just doesn't cut it.

Hybrid Refined (second picture)...that's pretty good. Something about the towers should be done to make them a little more graceful though. There are too many cables...but what could they do about that?

The two cable stayed bridges, 2 and 4 pier (bottom two pictures)...the 4 pier is ugly. The 2 pier is maybe o.k., but its towers still look like some kind of weird stargate. Is this really something that Portland should have to live with for decades to come?

If I understand things correctly, the height difference between the bridges pictured is just in the tower height. The decks are all roughly the same height. Except for being more of an aircraft hazard, that should be no big deal, especially considering the interest some people have in seeing tall structures such skyscrapers and smart-towers in the city.

[ac]

brian

bridges are appreciated from a distance or from within...either way, we're experiencing the structure as an object (from outside and usually from a distance beyond that of touching or a speed above walking)

so little jpegs are strangely appropriate to the scale

i can concede that they are, in fact, different from each other, but, for a project that can't afford a fine level of detail, it's the scale of the big idea that matters...in this one, the big ideas are basically the same

if the wave were done poorly (or "cheaply"...however you want to call it), it would still be the wave and offer a long distance image to the world around it

IMHO, these schemes offer the same view to all

[calvin]

I can't tell the difference - much ado about nothing. We either go for something that calatrava would do or what is pictured above. Everything else is arguing about peanuts.

The true test of whether you can tell the difference between the two is to imagine on the bridge poster in elevation and tell me that there will be a significant difference.

[Double J]

The so-called "refined cable stay" is pretty clunky. The V's are unresolved and would loom over the pedestrians and cyclists in a rather unrefined way.

Portland has to get more serious about design in its bridge construction projects, right now there is a lot of ad hoc futzing and it's almost like putting a red ball clown nose over the real nose... additive but incongruous.

Rosales had a decent ground up design for the Hybrid. I'm not blown away by it but I respect it.

Also, is height really the enemy here? If designed well the height is an asset and sends a strong symbolic statement about car-less transpo in Portland.

[pylon]

Rx'ed this in the mail the other day-

Portland-Milwaukee Light Rail Project Open House
Please join us for a progress report on the Portland-Milwaukee Light Rail Project, including an update regarding the planned Willamette River Bridge.
Thursday, June 11, 2009
4:30-6:30 p.m.
Willamette Room, first floor
2100 SW River Parkway, Portland

[pylon]

The V's are unresolved and would loom over the pedestrians and cyclists in a rather unrefined way.

...and could possibly create bird poo issues for those underneath?

[rob]

From a structural standpoint it is hard to see that a cable stay bridge would have to have the piers splayed out from top to bottom.