Flaws in the design led to cracks during post-tensioning of the first four concrete pontoons built for the replacement of the world's longest floating bridge in Seattle, says the Washington State Dept. of Transportation. The fix could cost tens or even hundreds of millions of dollars and potentially delay the $4.13-billion project.

In detailing a report from a blue-ribbon review panel, state Transportation Secretary Paula Hammond maintained that the pontoons—designed to a 75-year life cycle—are structurally sound. But the report revealed that five cracks up to .03 in. wide on Pontoon W could leak and therefore require retrofits.

Hammond made a public statement that "schedule pressure … drove many poor decisions" in a process filled with communication deficiencies and confusion regarding responsibilities. In the end, "inappropriate approval for post-tensioning tangent location change" was to blame, Hammond says, adding WSDOT will take appropriate disciplinary action.

Constructed in Aberdeen, Wash., by joint venture Kiewit-General under a design-build contract, the pontoons first showed cracking and spalling in May 2012 after initial post-tensioning of 60 strands in the 360-ft-long pontoons. All four pontoons in the first cycle were detensioned and repaired before float-out to Lake Washington last August. Additional in-water inspections, including the keel slabs, revealed additional lengthwise cracks, some reaching 10 ft.

The state convened a second expert panel in October 2012 to investigate the damages, leading to the agency's February admission that state engineers created a flawed design. All the pontoons need new transverse post-tensioning, including the first four structures already in the lake.

Replacement of the floating bridge over Lake Washington, connecting Seattle to points east, includes an even longer floating bridge, being constructed by joint venture Kiewit-General-Mason. At 7,710 ft, state Route 520 will cost about $954 million to replace, using 77 pontoons. Of that group, 54 are "flanker" pontoons that don't require post-tensioning, says Julie Meredith, SR 520 program director. But the spine of the bridge will take 21 longitudinal pontoons and two cross pontoons, all requiring post-tensioning for increased strength.

The first of six construction cycles included three longitudinal and one cross pontoon. The second phase, currently in progress, contains the same mix, with post-tensioning scheduled for April.

The panel report says, "The pontoon design did not adequately consider the effects of the post-tensioning layout, plus thermal and shrinkage effects," which can affect end walls' watertightness and service life. The bolt beam design was inadequate to resist post-tensioning splitting forces, the report adds. To close the cracks, all pontoons—whether new or constructed—will receive transverse post-tensioning for end walls and bolt beams.

Cycles three through six will simply add another step in the design phase to accommodate four transverse post-tensioning strands, says Larry Kyle, SR 520 engineering manager. Four pontoons in the lake and four in Aberdeen require two strands as a retrofit, he adds.

Crews will attach a steel coffer cell to the end of the pontoons in the water. The dewatered area will enable workers to conduct fixes in a dry environment. From there, crews will core-drill through pontoon walls and then string duct material and post-tensioning tendons through those holes, anchoring them side to side as close to the top and bottom of the bolt beam as needed. The transverse post-tensioning strands will squeeze together the longitudinal cracking. Also, injected epoxy will fill the cracks.

Coffer cell design is ongoing. Kyle says a "realistic schedule" would have the coffer cell ready by this summer. State officials and the contractor "don't know the final composition" of the sealant, Kyle says, adding fiberglass and carbon fiber are candidates. "It is an additional measure of safety to prevent any water intrusion if those cracks should open up due to wave action in the future," he says. "All of this is about preventing water from getting into the pontoons. The pontoons, even with cracks, are strong enough to withstand all the loads. This is all about sealing cracks and protecting reinforcing bars from water and corrosion."

The two cross pontoons are post-tensioned in the lengthwise and transverse directions at the bolt-beam location. Transverse post-tensioning will be added at the end walls.

Meredith anticipates that negotiations with the contractor on the cost for incorporating the new design elements into future pontoon construction will stretch until the end of the year. The cost for the in-water fixes is unknown. The total project, still more than $1 billion short in actual funding, had a $250-million contingency. About $200 million of the contingency remains, but the tricky western extension from the floating bridge to Interstate 5 hasn't commenced. Also, the Portage Bay bridge needs to be designed.

After interior spalling was found in cycle one in May, the first design change located the post-tensioning duct alignment for cycles two through six. Changes to the curvature of the ducts and shifting the anchorage six inches should eliminate spalling, Kyle explains. "We haven't post-tensioned the pontoons yet," he says. "We expect it will be fine."

For cracks less than .006 in. wide, considered within tolerance for concrete pontoons, crystalline waterproofing material—standard for WSDOT pontoon maintenance—will fill the cracks.