Schoharie Creek Thruway Bridge Collapse
This photo shows the 9 feet deep by 30 feet wide hole created by floodwater currents around pier three of the bridge. This scour hole should have been completely filled for the bridge to function correctly.
photo credit: U.S. Department of the Interior, U.S. Geological Survey (7)
Creative Commons license: Public Domain
You can view an excellent assemblage of photos, video of the collapse, and specific information here, on YouTube. Another short video showing the secondary collapse is available on the Wikipedia Schoharie Creek page here.
Location
The Schoharie creek bridge was located near Fort Hunter, New York, 40 miles Northwest of New York city. Fort Hunter is northwest of Albany in the Mohawk Valley. A satellite image of the bridge from google maps can be viewed here: Schoharie Creek Thruway bridge. The latitude and longitude at Fort Hunter are 42.942N and -74.285W. A general information page about Schoharie Creek can be found here, at Wikipedia.
History
The bridge was part of the Governor Thomas E. Dewey Thruway, where it crossed Schoharie creek. The Dewey Thruway is a 559 mile superhighway that crosses New York state. The Schoharie Creek bridge was one of several bridges constructed by the New York State Transit Authority as part of the highway project, and typically carried 15,500 cars and trucks each day (1). The bridge construction was completed in 1954, 33 years before the collapse in 1987. Six days after the collapse, a large part of the Mill Point Bridge collapsed, three miles upstream of the Schoharie Creek Bridge. Because the New York State Department of Transportation (NYSDOT) was concerned that the Mill Point Bridge foundations might also have been eroded by the flood, the bridge had been closed since the Schoharie collapse.
Details of the Collapse
The bridge collapsed the morning of April 5, 1987 during a near-record flood, after 33 years of use. There was six inches of rain and melting snow that combined to cause a fifty year flood. A fifty year flood has a 2% chance of occuring in any given year. Pier three collapsed, allowing spans three and four to fall into the flooded creek. One car and one tractor-trailer on the bridge fell into the creek. Three more cars were unable to stop in time and drove into the gap after the collapse, before the thruway could be closed off. Pier two and span two collapsed ninety minutes later; this second part of the collapse was captured on video. NYSDOT concluded that the wreckage from the initial collapse blocked the stream flow and diverted the flooding waters towards pier two with an increased velocity. This increased velocity and amount of water were able to topple the additional pier. All of the vehicle occupants died, with a total of ten fatalities. It took three weeks to recover nine bodies; one body was never recovered. Clearing the wreckage of the collapse was difficult due to the continued flooding.
About the bridge design
The bridge was composed of five simply supported spans connected by concrete pier support frames. The columns rested on lightly reinforced plinths, positioned on shallow reinforced spread footings. The spread footings were designed to be protected by dry riprap. The contract for the construction of the bridge was awarded Feruary 11, 1953, and the bridge opened after construction in the summer of 1954. The 1949 edition of the American Association of State Highway Officials "Standard Specifications for Highway Bridges" was used for the design of the bridge. Madigan-Hyland Consulting Engineers were contracted to complete the preliminary design. There were four piers to support the five simple spans: two on shallow footings in the creek, and two on the banks of the creek. There were also two abutments at either end of the bridge. The five spans were 100, 110, 120, 110, and 100 feet in length. On February 11, 1953, the construction contract was awarded to B. Perini and Sons, Inc.
Sequence of Events
In 1955, one year after the bridge opened, a 100 year flood occured. A 100-year flood has a 1% chance of occuring in any given year. The bridge survived that flood; however, the damage done to the bridge underwater and the subsequent changes in the creek bed as a result of that flood probably had an effect on the eventual collapse during the flood in 1987 (2). The collapse started when pier three fell, leading to the collapse of spans three and four. Ninety minutes later span two and pier two collapsed. Pier one and span one shifted two hours after the collapse of span two and pier two.
Cause of the Collapse
As-built plans showed that sheet piling had been left around the piers to protect them, but in fact it had been removed after constuction was completed (3). There were several problems that emerged shortly after the bridge construction was complete. The pier plinths began to form vertical cracks ranging form 1/8 to 3/16 inches wide and there was higher tensile stress in the plinths than had been expected; the cracks formed as the columns put bending stress on the plinths. Only the lower portions of the plinths had been reinforced because designers had thought this would be sufficient (2). The upper portions of the plinths were determined to need reinforcement, and in 1957 reinforcement was placed on the upper portions of the plinths to correct the cracking problem (2). After the 1955 100-year flood, the riprap was probably washed away; it was never replaced (4). The sheet piles seen in the as-built plans could have prevented the scour, but they were removed (2). In a 1980 maintenance contract, a nonengineer state employee deleted all reference to new stone riprap after viewing the site from shore and determining that it was unnecessary (3).
Two investigating teams determined that the cause of the collapse was excessive scour under pier three (4). The excessive scour was due to several factors. The depth of the shallow footings was not sufficient to protect from the amount of probable scour (5). Additionally, the foundation of pier three was placed on erodable soil. The area excavated for the pier should have been completely filled with riprap stone to resist scour, but instead was backfilled with erodable soil topped off with riprap (5). The stone riprap used was too light, and the protection, inspection, and maintenance of the riprap was inadequate (5).
The direct cause of the collapse was scour, aggavated by six factors (6). The flood was greater than the designers anticipated, and followed the 1955 flood that already disturbed the inadequate riprap. A curve in the river upstream directed faster water toward pier three. Drift materials were caught against the piers and pushed water downward toward the base of pier three. Berms were constructed upstream in 1963 and directed additional floodwaters under the bridge. An embankment west of the creek channel increased the speed of flood waters, and the Mohawk River downstream was set for winter conditions, affecting the flow gradient so that additional and faster currents were flowing under the bridge (6).
The indirect, human cause of the collapse was the failure to maintain the bridge riprap. The NTSB found that ambiguous construction plans and specifications, an inadequate inspection program, and inadequate oversight contributed to the cause of the collapse (4).
Collapse Outcomes
Investigators found that pier three had fallen into a scour hole 9 feet deep and 25-30 feet wide. Investigators determined that the riprap should have had weights from 1000 to 1500 pounds (5). The riprap used had 50% of stones over 300 pounds, and 50% between 100 and 300 pounds (5).
The severity of the collapse was affected by four factors of the bridge design and maintenance (6). The bridge bearings allowed the spans to lift or slide off the concrete piers. The simple spans were not redundant. The concrete piers were too static to allow for some frame movement. The plinth reinforcement stopped the hinge action of the plinth cracks; instead of sinking slowly into the scour hole, the plinth suddenly cracked and collapsed completely.
In the 1950s, nonredundant spans and bearings that allowed lifting/sliding were common (3). The NTSB learned that it is important to identify the critical features that can lead to a bridge collapse, then insure that those particular features are inspected correctly and frequently (3). In addition to designing bridges to resist the effects of scour, bridge designers and regulators learned that bridge specifications must include the selection of an appropriate critical storm/flood for the bridge design (3). Bridges need to regular inspections of the upper and lower structures, including underwater parts of the bridge (2). Finally, the bridge must be protected adequately from erosion around piers and abutments susceptible to scour (2).
References
1. Uhlig, M. A. (1987). 10 or more may be dead in bridge fall. New York Times. Retrieved May 5, 2008, from here.
2. The Collapse of the Schoharie Creek Bridge, summarized from Storey and Delatte, Lessons from the Collapse of the Schoharie Creek Bridge. Proceedings of the 3rd ASCE Forensics Congress, October 19-21, 2003, San Diego, CA.
3. Delatte, Norb. Maintenance and Management Lessons Learned from Bridge Collapses. Prepared for the Transportation Research Board Annual Meeting, 2007. Paper #07-2306. Abstract and retrieval information here.
4. National Transportation Safety Board (NTSB). Collapse of New York Thruway (I-90) bridge over the Schoharie Ckeek, near Amsterdam, New York, APril 5, 1987. Highway Accident Report: NTSB/HAR-88/02, Washington, D.C.
5. Thorton, C.H., Tomasetti, R.L., and Joseph, L.M. Lessons from Schoharie Creek. 1988. Civil Engineering, 58(5): 46-49.
6. Thorton-Tomasetti, P.C. Overview report investigation of New York state thruway Schoharie Creek bridge collapse. Prepared for New York State Disaster Preparedness Comission.
7. Scour at Schoharie Creek at Interstate 90 (West pier). 2002: http://ny.water.usgs.gov/projects/scour/fig5.html.