The new Sellwood Bridge will be a steel deck arch structure approximately 2,000 ft in length. Courtesy of Multnomah County 

A steel deck arch bridge will economically replace an aging Warren truss bridge in Multnomah County, and the existing bridge will be used as a detour while the new crossing is being constructed. 

September 11, 2012—When officials in Oregon’s Multnomah County decided to replace the Sellwood Bridge, which crosses the Willamette River near Portland, they knew they would need a functional design that would accommodate the 30,000 vehicles that make the crossing each day. They would also have to secure the landslide on the west side of the river, provide adequate space for pedestrians and bicyclists, and ensure that the design harmonized with the landscape. But they also knew that meeting those needs and aligning the project with their budget weren’t going to be easy. The project’s design and construction team is meeting the challenge by pursuing a cost-effective bridge type and integrating a slate of cost-saving measures that include moving the existing bridge and using it as a detour while the new crossing is being constructed.

Designed by the renowned bridge engineer Gustav Lindenthal, Sellwood Bridge opened to traffic on December 15, 1925. The Warren truss bridge has a steel deck, and its four continuous main spans over the river total 1,200 ft. The approach spans on the east and west ends total approximately 750 ft. The busiest bridge of its type in the state, the structure is 32 ft wide and has two 12 ft wide travel lanes and a 4 ft wide sidewalk that is shared by bicyclists and pedestrians. Ever since the bridge opened, it has been threatened by an active landslide along the west side of the river near the interchange with Route 43. The land has moved more than 3 ft over the years, requiring maintenance and major rehabilitation to the interchange and the approaches to the bridge. In 2006 the bridge was assigned a 10-ton load limit and given a sufficiency rating of 2 (out of 100) as a result of the effects of the shifting soil, says Mike Baker, P.E., M.ASCE, a vice president of David Evans and Associates, Inc., an engineering firm based in Portland. Baker is serving as the project manager for the owner's representative on the Sellwood Bridge project.

County officials considered rehabilitating the bridge, but the structure’s deficiencies were deemed so significant that replacement was the only practical option. They hired T.Y. Lin International Group, a multidisciplinary engineering services firm headquartered in San Francisco, as the prime contractor and CH2M HILL, an international engineering and construction firm based in Englewood, Colorado, as the subcontractor on the replacement project. They also awarded a construction manager and general contractor contract to Slayden-Sundt, a joint venture of Slayden Construction Group, Inc., based in Stayton, Oregon, and Sundt Construction, Inc., based in Tempe, Arizona. 

Designed by the renowned engineer Gustav Lindenthal, the Sellwood Bridge, in Oregon’s Multnomah County, opened to traffic in 1925. From its very first days, an active landslide along the west bank of the Willamette River has threatened the bridge’s structural stability. Courtesy of Multnomah County

Working with the project team, county leaders studied several bridge types for the new crossing, including a concrete deck arch and a concrete box girder. They eventually selected what was seen as the best value: a steel deck arch bridge. Some residents “would have liked some arches over the top of the bridge, [but] that was just a little out of our price range,” says Deborah Kafoury, a Multnomah County commissioner whose district includes the bridge. “The steel deck arch . . . has some artistic qualities to it. It’s visually appealing, but it was definitely on the lower end of the price range than some of the other designs we looked at.” Using steel rather than concrete to construct the bridge could save the county an estimated $4 million, Baker says. “That was planning-assessment level potential savings, but it was sufficient for us to solidify the choice of steel as the primary material of the bridge,” he says.

The new bridge will be approximately 2,000 ft long and, like its predecessor, will carry two lanes of traffic. However, it will be much wider. Its width will be 60 ft, except at the west end, where it will be 90 ft to accommodate turn lanes onto Route 43. In addition to two 12 ft wide traffic lanes, the bridge will have two 12 ft wide sidewalks and two 6 ½ ft wide bicycle paths to accommodate one of the most physically active communities in the nation. The bicycle paths will double as shoulders. The structure will also have four belvederes, two on the north side and two on the south, where pedestrians and bicyclists will be able to gather and enjoy the view of the river. “More than half of the bridge deck is given over to pedestrians and cyclists,” says Eric Lindebak, AIA, a principal with Safdie Rabines Architects, based in San Diego. He adds that the amount of space reserved for bicyclists and pedestrians is a significant change from the present bridge. “The existing bridge has maybe a three or four foot [wide] sidewalk,” he says. “It’s kind of like walking the gauntlet right now going across that river with the cars. It’s very narrow.” The replacement is also being designed to accommodate a future streetcar line that would extend down the middle of the structure. 

A number of retaining walls are being installed to stabilize the landslide on the west side of the Willamette River. Courtesy of Multnomah County

Although the bridge is in the final stages of design, construction began in December 2011 on one of the two contracts covering the early work. That contract was for the construction of the temporary piers that will support the bridge once it is slid over to serve as the detour. The second contract for early work is under way to secure the landslide. A variety of retaining wall types, from tied-back soldier piles and rock bolts to soil nails and conventional cantilever soldier piles, are being constructed as part of that effort. As a supplement, multiple tiers of pretensioned tieback anchors and a “wall” of 6 ft diameter cast-in-place drilled shafts will help stabilize the landslide. Tiebacks will be drilled on both sides of the footing on the west shore, descending approximately 160 ft. These tiebacks will provide a capacity of roughly 200 kips each. “There’s a Fort Knox–like structure buried underneath the ground that the public will never see that is being used to repress that landslide,” Baker says. “We’re spending almost thirteen million dollars just on addressing the landslide.”

The bridge will be supported by two piers in the river and a pier on each riverbank, all founded on 10 ft diameter concrete drilled shafts. The foundation shafts for the river piers, some more than 170 ft in length, will be drilled in the wet. Once the shafts are in place, the contractor will build a bathtub cofferdam around the shafts and then dewater those areas so that the piers can be constructed in the dry. The shafts will serve as anchors for the cofferdams, helping to control uplift. The team planned to use conventional cofferdams to construct the new piers, but when the temporary piles were driven for the detour bridge, the contractor hit large boulders, tree trunks, and other debris at the bottom of the river. As a result, the team decided to use bathtub cofferdams in constructing the new piers to avoid the need for the sheet piles required by traditional cofferdams, says Mike Lopez, P.E., S.E., a senior bridge engineer with T.Y. Lin International Group. “The pile driving was a lot more painful than we anticipated it being, so this cofferdam bathtub . . . gained a lot more appeal at that point,” he says. “That’s going to save a lot of time and risk . . . without having to build a conventional cofferdam that involves vibrating sheet piles into the river to a fairly great depth and then excavating riverbed material out to construct a footing and dewatering the enclosure.” 

The new Sellwood Bridge will have two 12 ft wide traffic lanes, two 12 ft wide sidewalks, and two 6 ½ ft wide bicycle paths, which will also serve as shoulders for the traffic lanes. Courtesy of Multnomah County

The bridge will have two main spans and a total length of 1,275 ft, and seven approach spans—two at the west end and five at the east—will add approximately 700 ft. The superstructure on the main spans will be concrete decks on longitudinal steel girders. The steel girders will frame into steel and concrete cap beams at the arches’ spandrel columns. Each arch will have four spandrel columns and two flanking bents. The approach spans will be supported by seven piers founded on drilled shafts. The superstructure on the approach spans will comprise conventional concrete bulb-I and bulb-T girders and concrete decks.

Expansion joints and expansion bearings will connect the approach spans to the main spans. In addition to the conventional approach spans, six other approach spans will run perpendicular to the bridge and parallel to Route 43. These spans will have a total length of roughly 600 ft. The bridge deck will be conventional reinforced concrete over the length of the structure, but it will be thicker than a typical deck to support the possible streetcar line. “We have designed an extrathick deck that can literally have a seven-inch-deep section removed, with streetcar tracks being added in the future,” says Eric Rau, P.E., a bridge engineer for T.Y. Lin International Group.

At every step of the project, the design and construction team is considering ways to save costs. These include using the boulders that were recovered from the river as construction material in another part of the project and investigating retaining wall options that could bring down the price tag. But the most significant cost-saving measure that the team is implementing involves using the existing bridge as a detour structure. “That idea alone, which was brought to us by the contractor, we’re assessing could be up to ten million dollars in savings and save several months of construction activity to accelerate the project,” Baker says. “Getting the bridge out of the way is going to allow the contractor to build the new bridge in one stage, versus the original two-stage approach.”  

The new bridge’s belvederes will enable pedestrians and bicyclists to gather and take in views of the Willamette River. Courtesy of Multnomah County

The temporary piers for the detour are already in place, and they line up with the existing piers. Crews will remove the existing bridge from its foundations and jack it up a few inches. They will then put the structure on a roller system and move it 40 to 60 ft north, where it will be skewed to connect with the new temporary east and west approaches. The team hopes to have the detour bridge in place by January 2013. That bridge “will stay in operation throughout the remainder of construction of the new bridge,” Lopez says. “Eventually, we’ll come back and deconstruct the detour bridge, take out the temporary foundations and piles in the river, and all we’ll be left with is the new bridge with its new foundations.”

While the team continues to explore additional money-saving measures, the project is expected to cost approximately $300 million. It is being funded by state and local dollars as well as by a $17.7-million Transportation Investment Generating Economic Recovery (TIGER) grant from the U.S. Department of Transportation. The project is scheduled for completion in 2016. When it’s finished, the community will have a new crossing that is at once safe, aesthetically pleasing, pedestrian and bicycle friendly, and within budget, Baker says. “The community told us that the bridge should be about more than just moving people across the river,” he says. “It should provide an opportunity for people to experience the river and enjoy the views and really interact [with one another]. It’s part of the community; it’s not just a bridge.”