Featuring developments in Federal highway policies, programs, and research and technology.
|This magazine is an archived publication and may contain dated technical, contact, and link information.|
|Federal Highway Administration > Publications > Public Roads > Vol. 73 · No. 5 > The Positive Legacy of a Bridge Collapse|
Publication Number: FHWA-HRT-10-001
The Positive Legacy of a Bridge Collapse
by Roger Surdahl, Donald Miller, and Vicki Glenn
The ripple effect of a 1989 bridge failure is still felt today as engineers and safety experts continue to update specifications for the construction of temporary shoring.
On the sunny summer morning of August 31, 1989, Stephen Brown was driving his regular rush-hour commute from his home in Maryland to Washington, DC, where he worked in the White House Communications Agency. While his eyes watched the road, his mind was reviewing the tasks to be tackled in the day ahead.
As Brown drove the Baltimore-Washington Parkway, he saw movement ahead on the Maryland Route 198 bridge, which crosses over the parkway. Construction had been occurring on the bridge for months, and seeing workers moving around overhead was nothing new. But on that morning, workers suddenly were moving fast, trying to get off the bridge.
With his attention riveted on the scene he was approaching at 55 miles (88 kilometers) per hour, Brown witnessed a slow-motion collapse as the structure fell, with still-liquid concrete literally flowing onto the parkway. Standing on his brakes, the 23-year-old stopped his car and maneuvered to the left shoulder fewer than 100 yards (91 meters) from the collapse. The car in front of Brown was not so fortunate, and he heard and watched the vehicle plow into the falling debris.
Brown also saw workers hanging onto rebar, trying to keep from falling onto the roadway below. With the help of another traveler, Brown extricated the driver of the car that had crashed into the debris. The driver had severe head injuries, and they moved her to the side of the road to await help. Fortunately, there were no deaths. Other than the driver in front of Brown, there were only several minor injuries, even though tons of concrete and steel had fallen onto a heavily traveled roadway.
In response to the collapse and to prevent future bridge failures, the Federal Highway Administration (FHWA) launched a series of initiatives, and soon the American Association of State Highway and Transportation Officials (AASHTO), American Society of Civil Engineers (ASCE), and others joined the effort to learn as much as they could about the cause of the collapse. The agencies set out to develop procedures and specifications to improve construction of shoring and other temporary works, which were discovered to be the cause of the MD Route 198 bridge collapse.
In fact, the U.S. bridge construction industry is still addressing the effects of the collapse, two decades later. Spurred by another collapse — the 2007 failure of the I-35W bridge in Minneapolis, MN — AASHTO published interim revisions in 2008 to its specifications for temporary bridge works. ASCE plans to publish revisions in 2010 to its guidance. And the National Cooperative Highway Research Program (NCHRP) will reevaluate standards for designing, building, and inspecting temporary works, also for release in 2010.
Over the years, the purpose of such publications — whether by FHWA, AASHTO, ASCE, or NCHRP — has been to increase national awareness of the importance of building temporary structures that ensure the safety of bridges, the workers who build them, and the traveling public. Specifically, as a result of the bridge collapse on the Baltimore-Washington Parkway, the goal was to develop guidelines and standards, improve specifications, and publish a shoring handbook, where previously none existed.
Scenic Highway to the Nation's Capital
In the early 1920s, planners envisioned an access-controlled scenic byway between the Nation's capital and Baltimore, MD, which became the Baltimore-Washington Parkway. Construction of the road became possible in the mid-1930s with the New Deal response to the Great Depression, but World War II intervened, delaying construction. The Maryland State Highway Administration owns and operates the northern 12-mile (19-kilometer) section, for which construction began in 1947. The National Park Service (NPS) owns and operates the southern 19-mile (30-kilometer) scenic section, for which construction began in 1950. The parkway opened to the public in October 1954.
Over the years, commuters increased traffic congestion, and the parkway was showing its age. There had been piecemeal upgrades to the facility, but in 1983, NPS began planning to upgrade the parkway. One component of the NPS portion of the parkway upgrade was in response to a 1983 inspection recommending that the two existing bridges carrying MD Route 198 traffic over the parkway be replaced with four new bridges: two each over the north- and southbound lanes of the parkway. These bridges were used heavily, as the MD Route 198 bridge is a main exit from the parkway to Fort Meade and the National Security Agency.
FHWA's Eastern Federal Lands Highway Division (EFLHD) designed the four new bridges and managed construction for NPS. The designers of the NPS structure also had to ensure the use of specific aesthetic features and materials.
Construction began in May 1987. The plan was to construct the two westbound bridges across the north- and southbound parkway lanes, then demolish the existing MD Route 198 spans and build the other two new spans at the site of the original eastbound lane of MD Route 198.
EFLHD designed the bridges as cast-in-place, posttensioned, concrete box girder structures with the eastern abutment bearings fixed and the western abutment bearing areas allowing for expansion. For Bridge No. 4, the simple span between abutments measured 95 feet (29 meters). Posttensioning in each girder web that provided the longitudinal strength for the box girder would occur at completion of the girder and as soon as the design concrete strength was reached. The final construction plans showed each bridge with a 30-foot (9-meter) horizontal width and 14.5-foot (4.4-meter) vertical clearance. The construction specifications required complete support by shoring until posttensioning could occur.
Bridge Nos. 2 and 4 were adjacent over the parkway's southbound lanes. The structure's design of concrete box girders and masonry-faced abutments met the aesthetic requirements of NPS.
A Bridge Fails
If there ever is a "good" time for a bridge collapse, it is late August in the Washington, DC, metropolitan area — many employees are on vacation, so fewer commuters are using the roads. A United States Park Police officer told The Washington Post shortly after the collapse that on the day the bridge failed, "[traffic] was probably 20 percent lighter than usual. It is common for cars to be stopped bumper-to-bumper during rush hour. There must have been a slight break in the traffic. It's a coincidence; a minute later or a minute earlier, and things could have been a lot different."
The bridge collapsed at 6:50 a.m. At 2:00 a.m., workers had begun placing concrete for the deck slab. Accounts by workers and inspectors confirmed the placement was going well, and shortly before 6:00 a.m. workers had placed about 120 cubic yards (92 cubic meters) of concrete, out of a total 160 cubic yards (122 cubic meters). The crews had finished more than half the 100-foot (31-meter)-long overpass deck.
About 15 to 20 workers were on the bridge at the time. Several reported hearing a breaking sound or feeling the bridge vibrate. Others said the sound was like a trailer being pulled under the bridge, and they looked to see whether a truck was passing underneath (commercial vehicles are prohibited on the parkway). Five to 10 seconds later, the center of the span abruptly collapsed, dropping fresh concrete, steel beams, wood, rubble, and workers onto the road more than 20 feet (6 meters) below.
"The bridge just vibrated for a second, settled back to normal, and all of a sudden it collapsed and people started falling," a 34-year-old worker told The Washington Post. He had plunged backward into a pile of wet concrete. "Fortunately, concrete, when it's wet, makes a good cushion," he said. "If it had been hard, it would have been a different story."
Another worker told how he landed on his knees in the rubble on the parkway and looked up to see the 1-ton (0.9-metric-ton) concrete smoothing machine looming above his and others' heads. "I just kept hoping that it didn't fall down on us," he said. (It didn't.)
As Stephen Brown's testimony confirms, drivers also had narrow escapes. Incident reports from the U.S. Park Police and followup interviews recount harrowing tales. A driver traveling south on the parkway saw the bridge begin to collapse. Debris struck the front windshield of his vehicle, but he was able to drive to safety. Another driver was in the last car to emerge from under the bridge. She saw the bridge collapse in her mirror. She reported having driven with a "pack" of other vehicles, all of which cleared the bridge and continued their commute, apparently oblivious to what was happening behind them.
U.S. Park Police and NPS personnel arrived within 10 minutes of the collapse, secured the scene, and redirected traffic as construction personnel, law enforcement, and fire and rescue units responded. Remarkably, the bridge collapse injured just 9 construction workers and 14 commuters.
By shortly after 7:00 a.m., area radio and television newscasts were broadcasting special reports to alert listeners and viewers of the situation.
Immediately, construction personnel and others involved in monitoring the concrete placement began to account for construction workers and motorists. The bridge had collapsed in the middle, forming a "V" as the crumpled structure hung from the abutments at either end.
Tense hours passed as inspectors and rescue personnel surveyed the scene and carefully began to remove debris and other material to determine that no people or vehicles were buried under the mountain of rubble. One worker had fallen onto the parkway during the collapse but immediately began to search the area beneath the bridge for vehicles or other workers. Ironically, he was the only person "lost," but the searchers soon accounted for him.
In late June 1989, Scott Wallace had become the FHWA project engineer overseeing the construction project. On August 31 — the day of the collapse — Wallace arrived onsite early to monitor the day's activities, including the concrete placement that began at 2:00 a.m. After hours of dividing his time between his office trailer and the bridge, he received a radio call that Bridge No. 4 had collapsed. He hurried to the site, which he would later describe as "bedlam."
For 3 days Wallace worked around the clock to ensure the cleanup went smoothly and that nothing prevented investigators from securing samples of intact shoring components and other construction material that might help identify the cause of the collapse. He later participated in tests on screw jacks and scaffolding conducted by FHWA at the Turner-Fairbank Highway Research Center (TFHRC).
"There were so many rumors — most wrong — circulating about the collapse and its cause," Wallace says. "There were also investigators looking into every aspect of the project. Essentially, I worked to make certain that events or processes [that were] questioned were correct and accurately conveyed, which they sometimes weren't."
FHWA Responds to Bridge Collapse
FHWA moved quickly to determine the cause of the bridge collapse and identify appropriate responses to improve bridge safety during construction. The agency established an investigation with three elements: It assigned responsibility to EFLHD to investigate the failure and report to the FHWA Administrator; hired T.Y. Lin International, a private consultant, to conduct an independent investigation and report the cause of the collapse to the Administrator; and established a Board of Review to evaluate all aspects of the investigations. The review board thus evaluated the EFLHD and consultant reports regarding the basic cause of the failure and recommended actions to prevent future collapses.
The result was the Board of Review's December 1989 Report of the Investigation into the Collapse of the Route 198 Baltimore-Washington Parkway Bridge (FHWA-PR-90-001), which includes findings from the EFLHD investigation, the independent consultant's investigation, and results of screw jack and scaffolding tests conducted at TFHRC. The report evaluated nine possible causes: severe ground vibrations, failure of stay-in-place deck forms, the concrete placement sequence, vehicle damage to shoring, movement of the shoring system from Bridge No. 2 to Bridge No. 4, failure of shoring foundation slabs, use of hardwood blocking, dynamic loading by the concrete finishing machine, and shoring failure.
During construction, but prior to the collapse, Tom McFadden, then NPS superintendent of Catoctin Mountain Park, Greenbelt Park, and the Baltimore-Washington Parkway, had voiced concern about ground vibration. "I recall sitting on the Bridge No. 4 headwall and watching two heavy trucks stop on Bridge No. 2," McFadden says. "Then as both trucks' engines revved and they began to inch forward, I felt the ground and headwall vibrate. It was a fluid feeling, like jelly."
Early in the inquiry, however, investigators focused on the shoring system. Timber framework on steel longitudinal support beams, which were supported by metal shoring towers, supported the superstructure construction on Bridge No. 4. The shoring consisted of three simple span sections: one 35-foot (11-meter) roadway span and two 28-foot (8.5-meter) abutment spans.
The shoring system had several changes from the original shop drawings. (1) The positions of the longitudinal support beams differed in relation to the concrete girder webs. (2) The leveling beams needed additional hardwood blocking underneath. As noted, workers feared a truck had hit the bridge; the original 12-foot (3.7-meter) clearance height was raised to 14.5 feet (4.4-meter) after an illegal oversize and over-height truck struck the shoring during construction of Bridge No. 1. The additional hardwood blocking was used to gain the needed clearance. With EFLHD's approval, the contractor (3) redesigned the deck overhang support system, (4) made changes to the reinforced concrete foundation slabs, and (5) changed the type of hanger rod assembly.
Following the collapse of Bridge No. 4, inspectors examining debris noted that many hanger beams were still intact even after sustaining the impact load from the collapse. They verified that the deck framework did not fail, and the positioning of the longitudinal support beams did not cause the collapse.
Inspectors examined the metal shoring system and confirmed it did not match the approved shoring drawings. After reviewing literature with the letterhead of the company supplying the shoring equipment, investigators asked representatives to visit the collapse site and confirm that their company had manufactured the shoring system components. Following that inspection, the representatives determined their equipment had not been used in the shoring system.
Further investigation revealed that two elements of the system did not conform to what EFLHD engineers had approved for use: The top screw jacks were not of the approved bearing capacity, and the cross-bracing was made of 1-inch (2.5-centimeter)-diameter tubing rather than 1.25-inch (3.2-centimeter)-diameter tubing. Most of the metal shoring elements did not bear a manufacturer's identification.
EFLHD designated screw jacks for a 25-kip (25,000-pound, or 11,375-kilogram) top and bottom shoring frame. But the screw jacks found at the collapse site were for a 10- to 11-kip shoring or heavy-duty scaffolding system. The top screw jacks also did not have the proper adapters to maintain vertical alignment in the structural tubing. The screw jacks were from different manufacturers and were 1.5 inches (3.8 centimeters) and 1.4 inches (3.6 centimeters) in diameter, while the opening in the shoring frame tubing was 2 square inches (12.9 square centimeters). Without the appropriate adapters, this created an unstable pin connection at the top screw jack-shoring interface.
In addition, investigators found that the top screw jacks were rusty, and many of the cross-bracing units had considerable rust and were heavily pitted. One cross-bracing member used a nail rather than the required bolt to connect the two cross-brace elements.
Investigations onsite and at TFHRC concluded that using improperly sized and poor-quality screw jacks and cross-bracing designed to be loaded at approximately 11 kips in the metal shoring elements — but loaded to approximately 25 kips — probably contributed to the Bridge No. 4 collapse.
After reviewing all findings and the position of the bridge following the collapse, investigators developed a probable failure sequence. The inadequate shoring system could not support the structure while workers were placing the deck slab. When the placement reached shoring tower 6 or 7, the bridge began to fail. Several eyewitnesses reported hearing a metal rattling noise just before the collapse, probably resulting from the cross-bracing as it failed and hit the metal shoring frame.
This failure redistributed the loads in the shoring system, and the remaining shoring towers in the first interior row of falsework beams on Bridge No. 4 (looking north on the parkway) failed almost instantaneously. The bridge failed at its midpoint because, before posttensioning, that is the weakest portion of a concrete girder section. As the bridge fell, it pulled its bottom flange off its bearing area, allowing the structure to fall to the ground. The bridge fell slowly, as evinced by drivers passing under the bridge as it began to fall and avoiding debris and emerging safely on the south side.
Board of Review Findings
The FHWA-appointed Board of Review concluded that it was reasonably certain the bridge collapse resulted from shoring failure at the second or third shoring tower (from the south) on the east side of the parkway. The board cited construction of the shoring towers out of compliance with plans approved by FHWA. Specifically, the contractor used 10-kip top screw jacks for the shoring towers rather than the specified 25-kip screw jacks shown on the plans. Moreover, the jacks used were in poor condition.
The board observed, "Other contributing factors included inadequate longitudinal strength in the overall falsework system, which was attributed to the smaller than required cross-bracing, a skewed shoring tower arrangement, excess blocking, a general lack of longitudinal cross support between the shoring tower members."
The Board of Review recommended measures to improve methods of shoring construction and approval to reduce the possibility of future shoring-related bridge problems. First, highway agencies should review and strengthen their specifications and construction procedures to better define the responsibilities of the material suppliers, contractors, and engineers. Second, agencies should provide a separate shoring design analysis for every bridge project. Moving shoring between bridges is an acceptable construction practice, but only when bridges are identical (which was not the case with the MD Route 198 bridges) and the shoring is thoroughly inspected for structural damage and plumbness when relocated, the board said. The inspection is to ensure that all members are in place and properly aligned and connected.
Contractors typically accept certificates for specification compliance from equipment manufacturers or providers. In the case of the MD Route 198 bridge, the contractor rented equipment and provided EFLHD the appropriate certification for elements of the approved shoring towers, but provided undersized jacks from various manufacturers. As its third recommendation, the board said highway agencies should — to ensure shoring systems are designed and constructed within manufacturers' design criteria, specifications, and load test data — require all shoring design submittals to be formally signed and sealed by the contractor's registered professional engineer, who will certify prior to placing loads that the shoring system has been assembled in conformity with approved drawings.
Fourth, the review board rec-ommended that highway agencies design each shoring system to handle all vertical and horizontal loading, and to have enough redundancy to prevent failure of the entire system. The agencies should consider vertical loading and differential settlement forces, plus lateral and longitudinal forces. They also should consider unbalanced temporary loading caused by the placement sequence, the board said.
Fifth, for shoring installations placed adjacent to open public roads, highway agencies should include special design considerations to ensure the shoring systems are not disturbed by errant highway vehicles or vibration forces caused by passing vehicles.
Sixth, agencies should move quickly to preserve and document in-place failure and assign investigation responsibilities to qualified, impartial parties.
Finally, FHWA should research manufactured shoring assemblies with the intent to improve current specifications and create new guidelines for design review and field inspection, the board said. The agency also should collect technical information on the various horizontal and vertical forces that interact in shoring tower arrangements.
The recommendations formed the basis of the 1990-1991 work of the Scaffolding, Shoring, and Forming Task Group, convened by FHWA, and the 1993 publication of a technical advisory (T 5140.24) encompassing four FHWA publications. (See "1989 Bridge Collapse Fosters 1991 Temporary Works Research Program" on page 26.)
The Difference One Bridge Collapse Can Make
Following the MD Route 198 bridge collapse, FHWA determined there was a need to reassess specifications used to design, construct, and inspect temporary shoring systems used in highway bridge construction. The agency created the scaffolding task group to coordinate this effort. One of the group's first actions was to form the FHWA Bridge Temporary Works Research Program.
John F. Duntemann of Wiss, Janney, Elstner Associates of Northbrook, IL, was principal investigator for the task group's contract in 1991 to research and develop design specifications for temporary works and a construction handbook for temporary works. He recalls that many States had very little guidance in their standard specifications regarding design, construction, and inspection of temporary works used in bridge construction. About two-thirds of States had some kind of submittal requirement for the temporary works, while about half had more specific requirements for designing the temporary structures.
AASHTO adopted the results of the task group's effort and published its own Guide Design Specification for Bridge Temporary Works and Construction Handbook for Bridge Temporary Works in 1995. As a result, many States updated their standard specifications, including the requirement that any significant plan revisions must be designed and sealed by a registered professional engineer.
A parallel effort to improve the design of temporary works was also active at ASCE. In 1987, a group of ASCE member design and construction engineers formed an ad hoc advisory committee to develop design standards for temporary structures used in construction operations. ASCE formalized the committee in 1989. Private sector contractors did not support the effort at first, fearing further codification of construction processes could affect a company's competitive edge.
After 15 years of work, ASCE published Design Loads on Structures During Construction (SEI/ASCE 37-02), which was embraced by private contractors. The publication applies to buildings and bridges, and the purpose is to provide minimum design load requirements for partially completed structures and temporary structures during construction. The standard does not specify the responsible party for design of temporary structures. Industry practice, however, is that the methods of construction are the responsibility of the contractor. The ASCE standard includes provisions beyond those in the AASHTO publications.
A Catalyst for Improving Bridge Construction Safety
The safety initiative continues, much of it in response to improved practices and materials. ASCE will issue its revised standards for publication in 2010. At AASHTO, the T-4 Technical Committee of the Subcommittee on Bridges and Structures published interim revisions to the Guide Design Specifications for Bridge Temporary Works in 2008.
"The 1995 guide design specifications helped define falsework and framework — the States may have used the terms but defined or interpreted them differently," says Malcolm T. Kerley, chairman of the subcommittee and chief engineer for the Virginia Department of Transportation. "The subcommittee revised the specifications in response to the I-35W bridge collapse in Minneapolis in 2007. The revision reflects updates [primarily from ASCE and AASHTO] from the last publication."
Ken Hurst, State bridge engineer for the Kansas Department of Transportation, chairs the T-4 Technical Committee, which approved the update. "We realized how out-of-date the 1995 specifications references had become, and in this first round they were updated. In addition, there are minor updates to the specification and commentary. But this was truly an interim step, and we need to look to the future."
The interim step anticipates the proposed NCHRP project to reevaluate the current state of the practice for designing, constructing, and inspecting temporary works, according to Hurst. This project also will revise other AASHTO specification and guideline publications, as necessary. "Stability is a concern," Hurst notes. "Shoring usually doesn't fall down; it falls over. With the exception of Load and Resistance Factor Design [LRFD], which most State agencies have adopted, there has been no major reevaluation of current practices in almost 15 years. It's overdue."
State transportation agency compliance with the AASHTO and ASCE provisions, plus adoption of the LRFD standards, have improved how projects are conceived and administered by agencies and their contractors, according to Duntemann, who was principal investigator for the 1993 publications resulting from the work of the FHWA Scaffolding, Shoring, and Forming Task Group. Duntemann now is principal investigator on NCHRP Project 20-07, Task 245, Updating AASHTO Design and Construction Specifications for Temporary Works Used in Bridge Construction. The literature review and research related to bridge construction practices for bridge temporary works is complete. Task 2, preparing an updated version of the AASHTO Guide Design Specifications for Bridge Temporary Works, should be published late 2010.
Duntemann notes that information gained from the initial part of the current project underscores the need to continue to evaluate and reevaluate the standards. He cites the collapse of the bridge over the Baltimore-Washington Parkway 20 years ago as the original catalyst, but the continued occurrences of these collapses reinforce the need for further development. "There is no question that FHWA, AASHTO, and ASCE initiatives improved standards that translate to better design and construction of these systems," he says, "but we anticipate that the updated guides and specifications will further reduce bridge failures in the future."
Stephen Brown, now 43 years old and an electronics technician in Washington, DC, is also grateful for progress in bridge design and construction. "It's been 20 years, but I still pay attention with my eyes every time I drive under a highway bridge."
Roger Surdahl joined FHWA in 1987 after receiving a master's degree in civil engineering from Montana State University. He is a registered professional engineer in Colorado. As the technology delivery engineer for CFLHD, he brings a wide range of experience in highway materials, contract administration, and innovative solutions to transportation problems.
Donald Miller has worked for FHWA since 1974 and holds a bachelor of science degree in civil engineering from Virginia Polytechnic Institute and State University. He has been with EFLHD for the past 36 years and is currently the director for project delivery.
Vicki Glenn is a consulting writer and editor who has developed and managed numerous FHWA technology outreach projects since 1992.
For more information, contact Roger Surdahl at 720-963-3768 or firstname.lastname@example.org, Donald Miller at 703-404-6201 or email@example.com, or Vicki Glenn at 301-514-6263 or firstname.lastname@example.org.
Page Owner: Office of Corporate Research, Technology, and Innovation Management
Scheduled Update: Archive - No Update
Technical Issues: TFHRC.WebMaster@dot.gov