IE 11 is not supported. For an optimal experience visit our site on another browser.

Bridge collapse revealed holes in federal data

When the I-35W bridge collapsed a year ago in Minneapolis, federal officials requested an emergency inspection of every similar bridge in the nation. There was just one problem: No one knew how many there were.'s Bill Dedman reports.
Emergency personnel survey the remains of the I-35W bridge in Minneapolis
When officials combed a federal database for bridges of the same type as the collapsed I-35W bridge in Minneapolis, they found the records were "not as good as we thought," according to e-mails released this week to Cohen / Reuters
/ Source:

When the I-35W bridge collapsed a year ago in Minneapolis, federal officials immediately requested an emergency inspection of every similar bridge in the nation. There was just one problem: No one knew how many there were.

At first, officials thought there were 756 steel deck truss bridges like the one that fell. That's how many they found in the official federal database of bridges, the National Bridge Inventory, which gets its records from the states. Then state engineers found 32 more to add to the list.

But when states started the inspections, they found that 280 of the bridges weren't steel deck trusses at all — including 13 bridges made of wood timbers. Another 16 no longer existed; a bridge in Pennsylvania had been closed in 1982. Another 11 were private bridges, not subject to federal inspection. One in New Mexico was a pedestrian bridge. And a bridge in Pennsylvania had been double counted; federal officials had placed an identical ghost bridge in Maryland.

By the time the survey was finished, the count of bridges of the same type as the Minneapolis span was down to 479, or 277 fewer than initially reported, according to internal e-mails from the Federal Highway Administration received Thursday by under the Freedom of Information Act.

All the steel deck truss bridges that could be identified did eventually receive their special inspection, almost all by Feb. 14. Only five bridges showed new significant problems.

But what about any bridges that might have been miscoded in the other direction and were still lurking in the database as another type of bridge? The federal e-mails don't address that question. The e-mails do, however, acknowledge a general weakness in the national system for keeping track of bridges: The "data is not as good as we thought," Thomas D. Everett, team leader of the bridge program at the Federal Highway Administration, wrote in "Issues of Concern," a slide show for safety experts on Jan. 13.

The 500 e-mails released to show federal bridge officials were operating under great stress in the months after the Minnesota collapse killed 13 people on Aug. 1, 2007. Officials were dealing with two federal audits of the bridge inspection process, fending off pitches from vendors of safety equipment, dealing with inquiries from members of Congress, answering questions from for its series on bridge inspection delays, writing talking points for the boss, all while continuing to press states to live up to their responsibilities to inspect bridges on time.

A sampler:

Kentucky was threatened with loss of federal funds for not turning in its bridge data.

Iowa bridge engineers expressed frustration that they weren't certified to inspect bridges, considering that they were considered qualified enough to design them.

Ohio was trying something new to make sure bridges were thoroughly inspected: Hiding tokens or small items on bridges, requiring inspectors to find them in a sort of treasure hunt.

The city of Los Angeles was told it had to let state officials do all inspections, because its inspectors weren't doing them properly.

Mississippi lost federal aid funds in February for seven cities that failed to post or close unsafe bridges, a month after had reported on the reluctance of federal officials to withhold funds when states didn't meet inspection deadlines.

Ohio was not assigning load limits to all bridges.

North Dakota was waiting four years to inspect fracture-critical bridges.

Hawaii officials, the farthest behind on inspections, asked for all training courses to be taught over the Internet; federal officials responded that you just can't teach someone how to inspect a bridge using the Internet alone.

And again in Kentucky, often the problem child, a state official asked how he was supposed to code in the database bridges that had been recommended for closure, but which remained open; the larger problem, federal officials replied: Why haven't the bridges been closed?

The National Bridge Inventory is used primarily as a tool for handing out federal highway money for repair of the worst bridges. But it's also the database used by the public, Congress and the media to get a snapshot of the condition and inspection history of the more than 600,000 bridges carrying traffic in the U.S.

"Everyone is 'data hungry,' particularly during times of crisis," Everett wrote. Recognizing the value of the data, the highway administration was trying to develop handheld computer access to the bridge database, for use by inspectors in the field; it shelved the project for lack of funds.

But in the "lessons learned" section of his presentation, Everett wrote that the Minneapolis experience showed that many pieces of information in the database were not being verified for accuracy by state and federal agencies. Many steps are being taken, he said, to improve the quality of the data.

Several state officials sent e-mails encouraging the federal bridge experts to keep their chin up.

"Hang in there, Tom," Terry Leatherwood, civil engineering manager for the state of Tennessee, wrote to Everett.

"Just remember that the the news media has the attention span of a fruit fly. So this will fade (as long as we don't lose another one)!"

The hundreds of e-mails released this week are being posted online at as .pdf files. Many of the e-mails have document files attached; look for those links at the top of the e-mails.