I-90 Pontoon (Floating) Bridges over Lake Washington at Seattle, WA

1940+1993 South Bridge, Lacey V. Murrow Memorial Bridge: (Bridge Hunter; HAER; Satellite)
1989 North Bridge, Homer M. Hadley Memorial Bridge: (Bridge Hunter; same satellite as above)

AARoads posted
The old Lake Washington Floating Bridge that carried US 10 at the time of this photo and later I-90. The bridge opened in 1940 and sank in 1990 during a storm.
[A second bridge was finished in 1989, and the original bridge was repaired by 1993.]

Hadley had help design concrete ships for the WWI Emergency Fleet. He had the idea in 1921 of building concrete barges with a deck to create a pontoon bridge over Lake Washington. He then scouted both Mercer Island and Seattle to find good locations for approaches to the bridge. His ideas were considered foolish until 1937 when he approached Murrow, director of the Washington Department of Highways "stating that he had found the most direct route for the bridge." Murrow started investigating his ideas and the more they studied them, the better they looked. A review committee "reached a unanimous opinion that the conditions at Lake Washington were uniquely favorable for a pontoon bridge. It had 'only a slight variation of its level, no current, no ice, no drift, and only limited reaches of open water in which waves could develop.' The engineers considered making the pontoons of steel or of wood as well as of reinforced concrete, but concluded that concrete pontoons, with necessary anchorages, was not only lowest in cost but also was superior because its greater mass would make the bridge it more stable. These studies, by the state highway department, included an experiment with a concrete barge anchored in Lake Washington, on the proposed route, from December 1937 to 31 March 1938. Various instruments on board recorded the tension in cables attached to the anchorages, daily weather conditions including wind direction and speed, the list of the barge, and the height of the waves. This taught the engineers much about the lake conditions and the possible behavior of pontoons." The pontoons of the resulting bridge were cabled to anchors, some of which were over 200' deep. Three types of anchors were used: deep soft mud, deep hard bottom, and shallow hard bottom.  (The reference has much more detail on the anchors.) [HAER-data]

Since Lake Washington is considered a navigable waterway, each end has a fixed 150' tied-arch truss over a navigation channel, a hinge, a transition span, another hinge, some transition pontoons (the deck is slanted to make a grade), an articulated joint, and then pontoons bolted together to the middle of the lake. At the middle, the is a pontoon that can slide out of the way to make a 200' wide, unlimited heighth, opening for boats that don't fit under an arch. Unfortunately, I have not been able to figure out from the text description how that movable span worked. I need to see some diagrams. Float valves in the transition pontoons change the level of water inside the pontoon to change the angle of the hinges to account for the 4' high variance of the lake's water level. [HAER-data]

Photo from HAER WASH,17-SEAT,13--4 from wa0182, cropped

A satellite caught a better view of the truss. We take tied arches for granted today. But in the late 1930s, when this one was designed, those arches had to be very innovative.

3D Satellite

In November 1990, while under re-construction, the original bridge sank because of a series of human errors and decisions. The process started because the bridge needed resurfacing and was to be widened by means of cantilevered additions in order to meet the necessary lane-width specifications of the Interstate Highway System. The Washington State Department of Transportation (WSDOT) decided to use hydrodemolition (high-pressure water) to remove unwanted material (the sidewalks on the bridge deck). Water from this hydrodemolition was considered contaminated under environmental law and could not be allowed to flow into Lake Washington. Engineers then analyzed the pontoons of the bridge, and realized that they were over-engineered and the water could be stored temporarily in the pontoons. The watertight doors for the pontoons were therefore removed.
A large storm on November 22–24 (the Thanksgiving holiday weekend), filled some of the pontoons with rain and lake water. On Saturday, November 24, workers noticed that the bridge was about to sink, and started pumping out some of the pontoons; on Sunday, November 25, a 2,790-foot (850 m) section of the bridge sank, dumping the contaminated water into the lake along with tons of bridge material. It sank when one pontoon filled and dragged the rest down, because they were cabled [actually, bolted] together and there was no way to separate the sections under load. No one was hurt or killed, since the bridge was closed for renovation and the sinking took some time. All of the sinking was captured on film and shown on live TV. The cost of the disaster was $69 million in damages. A dozen anchoring cables for the new Hadley bridge were severed, and it was closed for a short time afterward. Westbound traffic was allowed on Tuesday, and eastbound traffic was resumed in early December.
The disaster delayed the bridge's reopening by 14 months, to September 12, 1993.
[Wikipedia]

Because all of the pontoons were replaced, the bridge lost its historical registration.
(new window)

The 1990 maintenance was done after the new bridge was completed in 1989. In this Street View we see the old bridge is in service, and they are doing maintenance on the southern half of the new bridge. Normally those lanes are HOV reversible lanes. [1989 Bridge Hunter]

Street View on the original (south) bridge looking at the new one.

Bridges Now and Then posted
"Portal of Lake Washington Floating Bridge tunnel under construction, March 28, 1940" (Alfred G. Simmer/Museum of History & Industry, Seattle)

Street View, Aug 2021

A new tunnel and bridge now handles the other direction.

Street View, Oct 2022