Home > Latest

NTSB cites design errors for I-35W bridge collapse

The National Transportation Safety Board (NTSB) determined that the probable cause of the collapse of the I-35W Bridge in Minneapolis was inadequate load capacity because of a design error by Sverdrup & Parcel and Associates, Inc. (now part of Jacobs Engineering Group), of the gusset plates at the U10 nodes. The plates failed under a combination of substantial increases in the weight of the bridge that resulted from previous modifications, and traffic and concentrated construction loads on the bridge on the day of the accident.

Contributing to the design error was the failure of Sverdrup & Parcel’s quality control procedures to ensure that the appropriate main truss gusset plate calculations were performed for the I-35W Bridge and inadequate design review by federal and state transportation officials. Also contributing to the failure was the generally accepted practice among federal and state transportation officials of giving inadequate attention to gusset plates during inspections for conditions of distortion, such as bowing, and of excluding gusset plates in load rating analysis.

"We believe this thorough investigation should put to rest any speculation as to the root cause of this terrible accident and provide a roadmap for improvements to prevent future tragedies," said NTSB Acting Chairman Mark V. Rosenker. "We came to this conclusion only through exhaustive efforts to eliminate each potential area that might have caused or contributed to this accident. Bridge designers, builders, owners, and inspectors will never look at gusset plates quite the same again, and as a result, these critical connections in a bridge will receive the attention they deserve in the design process, in future inspections, and when bridge load rating analyses are performed."

About 6:05 p.m. Central Daylight Time on Wednesday, Aug. 1, 2007, the eight-lane, 1,907-foot-long I-35W highway bridge over the Mississippi River in Minneapolis experienced a catastrophic failure in the main span of the deck truss. As a result, 1,000 feet of the deck truss collapsed, with about 456 feet of the main span falling 108 feet into the 15-foot-deep river. A total of 111 vehicles were on the portion of the bridge that collapsed. Of these, 17 were recovered from the water. As a result of the bridge collapse, 13 people died, and 145 people were injured.

During its investigation, the Safety Board learned that 24 under-designed gusset plates, which were about half the thickness of properly sized gusset plates, escaped discovery in the original review process and were incorporated into the design and construction of the bridge.

On the day of the collapse, roadwork was underway on the I-35W Bridge, and four of the eight travel lanes (two outside lanes northbound and two inside lanes southbound) were closed to traffic. In the early afternoon, construction equipment and construction aggregates for making concrete were delivered and positioned in the two closed inside southbound lanes. The equipment and aggregates, which were being staged for a concrete pour of the southbound lanes, were positioned toward the south end of the center section of the deck truss portion of the bridge near node U10 and were in place by about 2:30 p.m.

Shortly after 6:00 p.m., a lateral instability at the upper end of the L9/U10W diagonal member led to the subsequent failure of the U10 node gusset plates on the center portion of the deck truss. Because the deck truss portion of the I-35W Bridge was considered non-load-path-redundant, the total collapse of the deck truss was unavoidable once the gusset plates at the U10 nodes failed.

The NTSB examined other possible collapse scenarios—such as corrosion damage found on the gusset plates at the L11 nodes and elsewhere, fracture of a floor truss, pre-existing cracking in the bridge deck truss or approach spans, temperature effects, and shifting of the piers—and found that none of these played a role in the accident.

As a result of its investigation, the NTSB made nine recommendations to the Federal Highway Administration and the American Association of State Highway and Transportation Officials dealing with improving bridge design review procedures, bridge inspection procedures, bridge inspection, training, and load-rating evaluations. Recommendations include developing and implementing a bridge design quality assurance/quality control program, and modifying approved bridge inspector training to address inspection techniques and conditions specific to gusset plates.

In response to the NTSB’s report, U.S. Secretary of Transportation Mary E. Peters said, "We are acting immediately to ensure that transportation officials nationwide understand and act on the Board’s findings. Many of the conclusions track with guidance we have previously issued to states over the past 14 months. Based on the Board’s findings and the need to ensure the structural integrity of gusset plates on bridges across the country, I am directing the Federal Highway Administration to work with the states to improve upon quality controls during the bridge design process. I am further directing the Federal Highway Administration to revise its training materials and inspection procedures to include specific guidance for reviewing gusset plates during all appropriate inspections."

Representative James L. Oberstar, chairman of the House Committee on Transportation and Infrastructure, predicted, "This will change all bridge inspections in the future. Inspectors will now have to go back to the design of the bridge itself, to review the engineering documents. They can no longer assume that the bridge was properly designed."

A synopsis of the NTSB’s report, including the probable cause, conclusions, and recommendations, is available online at www.ntsb.gov under the "Board Meetings" link. The board’s full report also will be available on the website.