The West Seattle Bridge, closed in March because of excessive cracking, might have been doomed since the day it opened in 1984.
City officials have listed several factors that could have contributed to the damage, including more and heavier buses and trucks, a seventh lane added years ago, a jammed rubber bearing that thwarts thermal expansion, and even the 2001 Nisqually earthquake.
But a hidden problem in the bridge’s original design might also be to blame.
A leading theory says the 220,000-ton bridge was gradually weakened by long-term shrinking of concrete within the twin girders that support the mainspan above the Duwamish Waterway. Local experts have pointed to an innate behavior of concrete known as “creep” that causes some bridges worldwide to sag by middle age. A crossing between islands in Palau even imploded after repairs.
“It is possible that the amount of creep that has occurred is greater than the designers, and broader engineering community, would have anticipated,” said Seattle structures director Matt Donahue, who ordered the West Seattle Bridge barricaded when cracks accelerated 2 feet in two weeks.
Now, contractors are racing to shore up the span, to preserve it long enough to determine if it can be repaired.
Volume loss from creep likely caused high-tension steel cables within the girders, which compress and strengthen the bridge, to slacken, University of Washington professor John Stanton said this spring. That in turn weakened the concrete and made it vulnerable to crack, he said.
In hindsight, consultant John Clark, who analyzed the cracks when the Seattle Department of Transportation (SDOT) discovered them in 2013, wishes he had pushed the city then to seal the cracks by re-tightening the mainspan with new steel.
“I should have been stronger seven years ago when I made the report,” said Clark, who worked as a design engineer on the bridge. He’s hopeful the span can be saved. “They left four 6-inch holes, in the event they felt they needed [more] post-tensioning. To me, it’s an active thing you can do.”
Clark also wishes the city had installed electronic devices in 2014 to allow round-the-clock crack monitoring
“I am surprised that as late as 2019, they hadn’t done more,” he said. SDOT is now attaching the monitors.
SDOT director Sam Zimbabwe, who arrived last year, has said the cracks didn’t pose a risk until recent months. Since 2014, the agency has applied epoxy, fastened gauges to measure crack width and increased inspections.
City officials say their own vigilance caught the more recent crack growth soon enough to protect the public.
Structural expert Evan Bentz, from the University of Toronto, advised this year that the cracks appeared consistent with “a collapse mechanism” but a collapse was not imminent.
Just in case, SDOT is writing emergency-alert plans to potentially evacuate the low bridge beneath the high-rise span.
Creep in concrete is driven by water-volume loss from dehydration, the compression of water molecules trapped in microscopic pores, or delayed chemical reactions.
Consultants are running computer models to estimate long-term creep for the bridge.
“However, right now our focus is on assessing and restoring the strength of the structure, not assessing creep deflections,” said Donahue, who added the bridge was built to standards that existed in the 1970s and 1980s.
West Seattle’s long, gently curved mainspan can’t support its own weight by concrete alone. To lend strength, the bridge is pre-stressed by hundreds of steel cables. Like a rubber band, the high-tension steel compresses the concrete to help it endure both the “dead load” of the bridge itself and the “live load” of traffic.
Thousands of pre-stressed concrete bridges have been erected in the nation since 1950, but this one was unusually ambitious, at 105 feet wide and 590 feet long in the mainspan. Concrete segments were poured 16 feet at a time during construction.
“For the West Coast, there haven’t been too many segmental bridges constructed,” Bruce Wasell, city project director, said during construction. “It’s even fairly rare in the United States. Probably within the last 10 years is the only time bridges have been built in this way.”
Design records estimated the concrete would shrink a couple of inches along the whole bridge. “It’s a difficult thing to measure,” Clark said.
Civil engineers in 1980 relied on models that assumed creep stops in 10 to 20 years after construction.
But in 2011, Northwestern University Professor Zdeněk Bažant — who devoted a half-century to investigating creep — examined data from 56 middle-aged bridges and found their pre-stressed concrete kept on compressing after 30, 40, 50 years, though at a decelerating rate.
“While a lifetime well in excess of 100 years is generally required in design, many of these bridges develop excessive deflections within 20 to 40 years. This may in turn cause cracking with corrosion, drainage problems, excessive vibrations, and car passenger discomfort. It may require either a bridge demolition or a retrofit with additional pre-stressing, which is a risky undertaking,” he wrote.
In 1996 a bridge in Palau, which slumped 5 feet, was tightened to restore its correct shape. Three months later, it collapsed at a preexisting crack, killing two people.
Creep also helped cause the Parrotts Ferry Bridge in central California to sag 22 inches, before the state saved it in 1990 by fastening a steel brace to the underside that resembles the arch support in a shoe.
While viewing West Seattle Bridge crack pictures, Bažant was struck by the 45-degree angle of the shear cracks, which he said indicates a loss of steel pre-stressing inside the girders.
Though the 36-year-old bridge contains 2 million pounds of tensioning steel, the primary pre-stressing cables go only through the central third of the high arc — which likely magnifies any problem caused by creep.
Cracks have formed symmetrically in both girders 112 feet shoreward from the bridge midpoint, just past where the steel ends are anchored. Clark’s 2014 study diagramed these transition zones between stronger and weaker segments, such as “joint 38,” east of the Delridge Way onramp.
“We believe that creep is likely a key reason the cracks at joint 38 continue to widen and extend,” Donahue said.
SDOT highlighted creep in an April 2019 work order for consulting engineers to recalculate the bridge’s strength. The root cause of cracking was “not yet known,” the contract said, but added, “Due to the nature in which the structure was constructed, it is known that load redistribution would have taken place due to creep and shrinkage of the superstructure.”
Ideally, any added post-tensioning steel should go from pier to pier, so it would compress the whole mainspan uniformly, Clark said last week.
Any time forces change, whether there is a stuck bearing, or cracks, or future modifications by SDOT, those changes will spark a new round of creep, said concrete scientist Neil Hawkins, who supervised the West Seattle Bridge design competition.
Donahue said “temporary post-tensioning” will likely be applied soon, with carbon fiber reinforcement, to stabilize the bridge.
Last week, the city announced the hiring of Wisconsin-based Kraemer North America to strengthen the girders from within by late summer. Then, SDOT must replace the stuck bearing on a Harbor Island pier, which blocks the bridge’s normal thermal movement.
If both those tasks succeed, then SDOT will be able to shore the bridge and attempt repairs next year.