Collapse Scene Investigation (CSI) Carmel River Bridge Failure - VZB
vzb.baw.de/publikationen.php?file=icse4/0/f_5.pdf
1. Collapse Scene Investigation (CSI). Carmel River Bridge Failure. Catherine M.C. Avila, P.E., D.WRE, MSCE, MBA. Member of ISSMGE, President, Avila and ...
1
Collapse Scene Investigation
(
CSI)
Carmel River Bridge Failure
Catherine M.C. Avila, P.E., D.WRE, MSCE, MBA
Member of ISSMGE, President, Avila and Associates Consulting Engineers, Inc.,
(San Francisco, CA, USA)
E-mail: cavila@avilaassociates.com
The Collapse of the 1995 Highway 1 Bridge in 1995 caused a 400 km detour to get around the collapsed
bridge. The clues from design and maintenance reports will be followed including channel bed degradation
and local scour estimates to determine the cause of the bridge failure and potential missed opportunities to
fix the bridge before it collapsed. The economic consequences of the bridge failure including lost
opportunity cost will be summarized.
Key Words :
Carmel River, Bridge Scour Failure, Detour Cost, Temporary Bridges
1.
F
OLLOWING THE
C
LUES
(1)
Setting the Scene
The project site is loca
ted just outside of the City
of Carmel located in Mont
erey County in Central
California as shown in
Figure 1
.
Figure 1
: Location Map within California
As shown in
Figure 2
, there have been three state
highway bridges at the site and a pedestrian bridge is
proposed.
Figure 2
: Past, Present and Future Bridges at the Site (basemap
from googleearth.com).
The earliest bridge with records available was
constructed in 1909. This 1909 timber bridge was
replaced in 1934 with a bridge upstream because the
timber piles and deck were
rotting and the bridge was
settling. The 1934 bridge was replaced in 1995 with a
707
F-5
Fourth International Conference on Scour and Erosion 2008
2
reinforced concrete box girder bridge upstream when
three spans of the 1934 bridge washed into the river
on March 10, 1995. A pedestrian bridge is currently
proposed upstream of the 1995 Route 1 Bridge as part
of a pedestrian and bicycle corridor planned by the
Transportation Agency of Monterey County
(TAMC).
The Carmel River is a central California coastal
stream that drains a 23.7 sq meters watershed to the
Pacific Ocean (Figure 3). The river has two dams on
its mainstem: the 25.9 m high San Clemente Dam
located at Rivermile (RM) 18.6 and the 42.1 m high
Los Padres Dam located at RM 23.5.
Figure 3
. Carmel River Basin (from Monterey Peninsula Water
Management District).
The gage upstream of the bridge (Carmel River
near Carmel, Gage #11143250) has been in place
since 1964. The highest recorded discharge was 453
cubic meters per second (cms) on March 10, 1995
when the bridge failed as shown in Figure 4.
Figure 4
: Discharges at the Carmel River Gage near Carmel
(note discharges in cubic feet per second)
The FEMA study estimated the 50-year discharge
to be 651 cms and 100-year discharge to be 824 cms
(FEMA, 1991). The March 10, 1995 discharge of 453
cms equates to approximately a 30-year discharge
assuming the above hydrology is valid. The
proposed pedestrian and bicycle bridge over the
Carmel River provides an excellent opportunity to see
if the forensic clues that
led to the March 10, 1995
Carmel River bridge failure can be used to enhance
the longevity and survivability of the new bridge
under a variety of flood events.
(2)
Clue #1: 1934 Bridge Design
The 1934 Carmel River Bridge was constructed
approximately 43 m upstream of a timber through
truss and trestle spans co
nstructed in 1909. The
bridge had four 20.1 m timber through trusses on
“mass concrete” on timber pi
les in the main channel
and eleven 6.7 m timber trestle spans on timber piles.
The old bridge was 155.4 m long and 5.5 m wide and
had a soffit elevation of 7.4 m. Note all elevations are
in National Geodetic Verti
cal Datum of 1929 which is
the datum of the as-built plans for the 1934 bridge.
The 1909 bridge was replaced because the piles
under the trestle and other parts of the bridge were
badly rotted out making the bridge structurally
unsafe. The north pier ha
d settled .3 m or more and
the trusses had been blocked up (to prevent more
settlement). The waterway
under the trestle portion
had been seriously encroached upon by a network of
bracing placed there not
long before. (State of
California, 1931).
The 1934 bridge was designed for a 75-year
discharge of 496 cms – only slightly larger than the
453 cms that destroyed the bridge. At the time of the
bridge design, there were no dams on the Carmel
River, and the “high-water”
was estimated to be at
elevation 6.1 m. The soffit elevation was
recommended at 6.7 m while another portion of the
report notes that the “elevation of high-water taken
from contour map on data by Mr. Mitchell, 22.0”
(State of California, 1931).
The bridge which was designed was made longer
(158.5 m instead of the proposed 146.3 m) than the
preliminary recommendations. In addition, a vertical
curve with a maximum soffit elevation of 7.9 m and
minimum soffit elevation of 7.6 m was added
provided as shown in Figure 5.
708
3
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
9.00
10.00
148+00 148+50 149+00 149+50 150+00
Bridge Deck Elevation
1930 Ground
1972 Ground
1993 Ground
1995 Ground
Soffit Elevation
Figure 5
: Ground Elevations and Profile of the 1934 Bridge
(meters)
The 1934 bridge was desi
gned assuming no cutting
down (degradation) and probable silting up, given its
proximity to the ocean. Interestingly, the proposed
materials for the bridge were gravel transported by
rail from Monterey as well as sand from the
streambed. The report noted an open commercial
sand pit 183 m upstream of the bridge. This may
have contributed to the degradation noted below.
(3)
Clue #2: Maintenance Records
The maintenance records for the 1934 bridge
(Caltrans, August 1995) in
dicate significant drift
caught on the piers between 1955 and 1986. In 1972,
the encroachment of a cultivated field encroaching
into the waterway between Abutment 1 and Pier 10
was detailed. This encroachment prompted the
inspector to decrease the waterway adequacy of the
bridge, and may have contributed to the amount of
water flowing under the bridge instead of out into the
historic floodplain. In 1958, structural damage of
Piers 11 and 12 was detailed, and the upstream pile of
Pier 6 was also knocked out of plumb. Later in 1977,
the 5
th
pile in the bent is noted as being cracked and
Pier 12 was the first pier to fail in 1955. The 1991
report notes that the bridge was placed in the
maintenance program for pile encasement of Piers 10,
11 and 12 (the bridge predated the Caltrans Scour
Countermeasure program which started in 1997).
(4)
Clue #3: Cross Sections
A cross section was taken on March 2, 1995, just 8
days before the bridge failed. According to the cross
section information (see Figure 5), the river degraded
approximately 1-meter between 1931 and 1995. The
degradation decreased the pile embedment from it
original 5 m at Pier 12 and 5.4 m at Pier 11 to 4 m at
Pier 12 and 4.4 m at Pier 11. The original estimated
pile penetration was 5.5 m according to the 1931
Preliminary Investigatio
ns Report (State of
California, 1931), which is
greater than the actual
penetration achieved in construction.
(5)
Scene Reconstruction: Hydraulic Model
A hydraulic model (HEC-RAS) set up by Balance
Hydrologics, Inc. was
utilized to estimate the
hydraulics at the bridge as it was in 1995. Although
the model utilizes the 2007
topography, if
the record
discharge of 453 cms is contained in the channel, it
provides an estimate of the water surface elevation
and velocity for the old Highway 1 Bridge as shown
in
Figure 6
.
Figure 6
: Output from HEC-RAS Hydraulic Model
(6)
Clue #4: Debris and Pressure Flow
Debris pummeling the bridge was caught on amateur
video late Friday afternoon just hours before the bridge
collapsed by a local reside
nt who commented that it
“felt like an earthquake.”
The documented debris
included debris rafts and large trees, which were over
30.5 m long with 1.8 m diameter root-balls. The water
was at or near the soffit of
the bridge with significant
debris catching on the piers. Debris lodged on a pier
can increase local scour at the pier by increasing the
pier width and deflecting a component of flow
downward. HEC-18 recommends estimating the scour
depth by assuming the pier width is larger than the
actual pier width and notes the problem is determining
the increase in pier width
to use in the pier scour
equation. According to HEC-18, limited studies of
pressure flow scour indicate that pier scour can increase
200 to 300% for a bridge under pressure flow
(Richardson et. al., 2001).
With relatively small piers
(0.46 m), the pier scour
is 1.1 m. This leaves 3 to 3.3 m of pile embedment. If
the pier width is assumed to double due to debris, the
pile embedment is decreased to 1.8 to 2.1 m.
Likewise, if the pier scour depth is doubled or tripled,
the pile embedment is dramatically decreased.
According to the video (MPWMD, 2008), Pier 12
was the first pier to fail with
0.3 to 0.5 m of settlement
showing as the bridge began to collapse.. Later video
shows the complete loss of
Span 12 and the tipping of
Span 11 into the water. By morning, Spans 10-12 had
disappeared into the river
with just a downstream gas
line spanning the 36.6 m opening.
709
4
(7)
Follow the Clues . . .
The four factors leading
to the bridge failure
presented above include: Clue #1: 1934 study
underestimating the discharge and potential
degradation; Clue #2: Waterway encroachment and
structural damage concentrating the flows and
possibly weakening the bridge; Clue #3: Channel bed
degradation decreasing the pile embedment; and Clue
#4: Debris and pressure flow putting stress on the
bridge. To sum up, the co
mbination of relatively
small pile embedment, pressure flow scour, debris
and impact loading from la
rge trees floating down the
river likely caused
the bridge failure.
2.
T
EMPORARY AND
P
ERMANENT
B
RIDGES
In order to open the route to traffic as quickly as
possible, a Bailey bridge which was stored at the
Caltrans Maintenance yard at the San Francisco
Oakland Bay Bridge was tr
ucked down from Oakland
to Carmel. The turnaround was fast: the 1934 bridge
failed late Friday March 10
th
. The Bailey bridge was
“designed” and loaded on trucks on Saturday, March
11
th
. The new bridge erection started on Sunday and
was completed Thursday, March 16
th
(see Figure 7).
Figure 7
: Looking downstream at Bailey Bridge Launching in
March, 1995 (photo by Steve Ng, Caltrans)
The new bridge was designed while the Bailey
bridge was under construction. The recommendations
were 0.9 m of freeboard over the estimated water
surface elevation during the 1995 storms with 30.5 m
spans. Work started on the permanent bridge on April
1, 1995. It was opened to traffic on May 5, 1995 and
was completed by September 1995 (the training walls
were installed after
traffic), as shown in Figure 8. The
bridge cost almost $3.5 million U.S. dollars (Caltrans,
September 1995).
Figure 8
: New Highway 1 Bridge with Pier Wall Construction
under way (photo by Author, 1995)
Since the bridge was constructed, significant
analysis and improvements to the channel have been
made. The County of Monterey is currently
completing a “restudy” of th
e FEMA floodplain. This
analysis shows a decrease in the water surface
elevation due to the remo
val of some levees and
improvement to the downstream lagoon. These
improvements are lowering the anticipated water
surface elevation as the channel uses more of its
historic floodplain. The proposed pedestrian bridge
was designed with 0.9 m of freeboard above the
revised 50-year discharge to allow debris to pass
under the bridge. 30.5 m long spans that line up with
the downstream Route 1 Bridge were also
recommended.
3.
E
CONOMIC
C
ONSEQUENCES
There is significant economic cost associated with
channel bed degradation and lateral channel
migration. This includes not
only the cost of repairing
or replacing the bridge, but also the lost opportunity
cost associated with significant detours and/or traffic
delays.
Due to the bridge wash-out
, the trip from Carmel
Highlands to Carmel was increased from 15 minutes
to 6.5 hours due to the 400 kilometer detour (San Jose
Mercury News, 1995). Some researchers have
attempted to estimate the lost opportunity cost
associated with detours. A 1986 study in Texas
estimated the value of time to be $8.00 per
vehicle-hour for drivers, $10.40 per vehicle-hour for
passenger vehicles (assume 1.3 persons per vehicle)
and $19.00 per vehicle-hour for trucks (McFarlan and
Chui, 1986). These costs are approximately 20%
higher when updated to 1996 dollars (Caltrans, 1996),
or $9.60/veh-hr for drivers, $12.48/veh-hr for
passenger vehicles, and $22.80/veh-hr for trucks.
Therefore, the cost to an individual driver (in 1995
U.S. dollars) to drive fr
om Carmel Highlands to
Carmel was increased from $2.40 to $62.40 when the
710
end quote from:
http://vzb.baw.de/publikationen.php?file=icse4/0/f_5.pdf
If you want to see all the pictures and details please go to the url right above this line.
No comments:
Post a Comment