To: The Members
There are an estimated 12 floating, large-scale pontoon bridges in operation
world-wide, with the majority located in the Pacific Northwestern United
States. These unusual types of bridges are composed of a series of watertight
concrete boxes (similar to a barge) that are bolted together to form a continuous
structure and held by cables anchored in concrete at the bottom of a river
or lake bed. The top surface forms the roadway for vehicular traffic. Typically
these bridges are found in locations where there are low fluctuations in
water levels, and where soft soil conditions and significant water depths
make the sinking of pilings cost-prohibitive.
Two such large-scale pontoon bridges in Washington have experienced failures
stemming from a combination of the under-design of the anchoring system,
obsolescence of the design criteria and lack of engineering experience with
pontoon bridge design. Poor maintenance may also have played a part in the
losses. Despite these two pontoon bridge failures the Washington Department
of Transportation (WDOT), and bridge designers in general, remain committed
to the design and have worked to refine it over many decades. This may signal
While the average underwriter may not frequently come across large-scale
pontoon bridges, smaller scale pontoon bridges can be found throughout the
country, including those that are used to cross shallow, marshy waterways
or as temporary crossings while permanent stationary bridges are under construction
or rehabilitation. With temporary pontoon bridge applications, economics
rather than geography is often the primary factor in their usage. In addition,
the privatization of such bridges is an emerging trend that may also result
in new underwriting opportunities.
This bulletin will examine two infamous Washington pontoon bridge losses,
and also discuss underwriting issues related to completed pontoon bridges
as well as builders' risk involving rehabbing and replacement construction
of these bridges.
Hood Canal Bridge -- 1979
The Hood Canal Floating Bridge, which connected Kitsap Peninsula with the
Olympic Peninsula west of Seattle, was built in 1961 at a cost of $24.6
million. The bridge was 6,520 feet long and consisted of 50 floating concrete
pontoons, each 360 feet long and weighing 500 tons. At the time the bridge
was built, two contracting firms issued independent statements questioning
its design, and noting that maximum wind, wave and current conditions at
the site could produce forces in excess of the design specifications.
The firms' fears proved to be well-founded when a vicious winter storm hit
the area in February, 1979. With winds registering a sustained 87 miles
per hour (mph) and gusting to 100 mph, the waves and currents produced forces
that resulted in destructive stresses. The storm's intensity was the highest
ever recorded at the bridge site, and caused damage to half of the pontoons.
Water rushed into hatches and bolt holes, and 13 pontoons sunk to the bottom
of the 300 foot deep canal. The amount of excess water in the pontoons was
too much stress for the design.
Repairs were under way by 1982. The WDOT decided that the depth of the canal
bottom would make rebuilding as a conventional bridge both costly and difficult.
The western section of the bridge (3,800 feet long) had to be completely
replaced, while the eastern section required repairs of only some of the
The new bridge length would total 7,131 feet. The rebuilt bridge would be
one of only five floating bridges in the world at that time, and the only
one to be located over tidal waters. The new bridge section would be 10
feet wider, four feet deeper and 50% stronger than its predecessor. Anchors and pontoons would be heavier, and
the steel cables would be thicker to resist the high winds, strong currents
and heavy seas which devastated the original bridge. The replacement of
the western section would be the main focus of reconstruction, and the contract
was bid for $60 million.
One of the most intriguing construction features of the replacement bridge
project was the construction of a barge system that resembled a catamaran.
The barge system was designed to provide stability for lowering the bridge
anchors in place. The crane and winching systems were also custom built.
The pontoons were built 600 miles away and were towed to the construction
Lacey V. Murrow Bridge -- 1990
The Lacey V. Murrow Floating Bridge, the first major floating concrete bridge
in the world, was built in 1940 and crossed Lake Washington -- connecting
Seattle and Bellevue on Mercer Island. The floating portion of the bridge
consisted of 25 pontoon sections anchored in 200 feet of water by steel
cables. Each pontoon was a hollow, airtight steel box encased in concrete,
measuring approximately 350 feet in length, 60 feet wide and 15 feet deep,
and weighing approximately 4,900 tons. The floating pontoons were bolted
together to form one continuous structure, with the top surface utilized
as a roadway for commercial and passenger traffic.
At the time of loss, the Murrow Bridge had been undergoing extensive rehabilitation
(estimated contract cost of $35 million) and had been closed to traffic
for approximately 17 months. The sinking began the morning of November 25,
1990, with pontoon "A5," centrally located in the bridge. As the
pontoon settled into the water, additional water was taken on through open
maintenance hatches on the north side of the bridge. Pontoon A5 sank, dragging
additional sections of the bridge underwater. Within 8 hours, eight pontoons
were at the bottom of Lake Washington. The remaining pontoons were towed
to temporary storage sites along the lake in an effort to mitigate any additional
Considerable controversy arose over the cause of loss -- with the contractor
and the WDOT each pointing fingers at the other. After extensive investigations
spearheaded by both sides, the dispute was ultimately resolved through a
The WDOT findings indicated that the primary cause of sinking was water
accumulating in the bridge pontoons during the renovation work. The sources
of the water, in order of volume, were from the hydro demolition machines
used to remove concrete from the pontoons, rainwater that fell on the bridge,
and wave splash water. Environmental restrictions mandated that water used
in the demolition process be contained in the pontoons and carted off-site
for disposal, rather than discharged into the Lake. Although the plan was
to limit the accumulation of water stored in the bridge at all times, significant
hydro demolition work was conducted during this period. On November 22-23,
a severe wind storm occurred, and on November 23-34 heavy rains (4-5 inches)
fell. These circumstances, together with the age and condition of the bridge,
were held by the WDOT to contribute to the failure.
The contractor's investigation revealed that the probable cause of failure
was a progressive bond slip at lapped splices in the bottom slab of pontoon
A5, due to fatigue in bond. The failure resulted in the widening of existing
cracks, allowing water to flow into the pontoon. The deterioration was the
result of cumulative damage caused by wave-induced stress reversals experienced
over the 50 year life of the bridge.
While the original Murrow bridge had been designed to carry about 2,500
vehicles per day, it supported 105,000 vehicles per day just prior to its
close for rehabilitation. While the bridge design remains practical its
structural integrity was diminished. The replacement bridge, which cost
an estimated $80 million, utilized the same design concept while incorporating
the following engineering improvements:
- Prestressed high performance silica fume concrete
for pontoon fabrication
- Larger, heavier and thicker walled pontoons with
heavier rebar enforcement
- Post tensioning of pontoons
- "Egg carton" pontoon design for airtight
Additional measures taken to prevent water accumulation
in the pontoons included patrolling by maintenance boats outfitted with
pumps for water removal, utilization of scuppers along the bridge to drain
off water, and installation of high water level sensors and automatic bilge
pumps on each pontoon.
Underwriters need to know what wind and flood loads a pontoon bridge design
can handle. As a minimum standard, the design should meet at least a 100-year
event (e.g., 100-year storm, 100-year flood).
There are other important considerations for underwriters,
- How have the values been established? The potential
repair costs of older pontoon bridges are often several times the original
cost of the structure. How are the values updated? Are appraisals available?
- The business interruption (or loss of use) from
the loss of a bridge can be substantial. Are values reported accurately?
Are there alternatives routes that could mitigate a business interruption
- Does the insured have a comprehensive maintenance
program in place? Are records available for review? Are inspections conducted
annually? What is the frequency of underwater inspection of the cables/anchors?
- For new or rehabilitation of existing bridges,
underwriters need to know the qualifications of the bridge design team.
Do they have experience with this type of bridge? What are the qualifications
of the General Contractor and the key subcontractors?
Note: For underwriting concerns relevant
to other bridge types, see the 1990 IMUA paper entitled "Bridges, What
Can We Learn?".