A consultant urged new plates for the aging I-35W bridge, the same plates that are now the focus of a federal inquiry, records show.
In the year before the Interstate 35W bridge collapse, the engineering firm that studied the structural safety of the bridge found itself at odds with state transportation officials as it pushed a plan to reinforce the aging structure.
According to newly released documents from the Minnesota Department of Transportation, consultant URS Inc. played a diminishing role in decisions about the bridge even as it prepared its final report on its maintenance.
The records also show that URS was puzzled when MnDOT suddenly cast doubt on the consultant's $2 million plan to strengthen the bridge with steel plates.
The documents -- a mix of 13,500 e-mails and other communication from 2003 through 2007 -- offer the most complete look yet at the partnership between MnDOT and the San Francisco-based consultant, which had spent three years doing computer modeling and analysis on the 40-year-old bridge.
The Star Tribune obtained the documents through a state data practices act request.
In the days after the collapse, MnDOT said state officials and URS had mutually decided to explore other options instead of bolting steel plates to bridge sections at risk of cracking. But the documents indicate the two sides had a substantive disagreement over how to proceed.
Last December, after pushing for months to replate the bridge, URS abruptly moved closer to MnDOT's position of finding a less intensive way to ensure the bridge's structural integrity. Three weeks after a top URS official had reiterated that the chance of a bridge truss failure "should be significantly reduced" by the replating, the same official suddenly e-mailed a colleague that he no longer thought the replating was necessary.
Even the agreed-upon solution -- closer inspections -- did not proceed as URS expected. On July 19, two weeks before the bridge collapsed, URS official Don Flemming was caught by surprise when MnDOT inspected the bridge without the firm. "I saw them on the bridge, and asked if we were not going to get involved," wrote Flemming, who for 14 years had been the state's bridge engineer before being hired by URS.
MnDOT spokesperson Lucy Kender said in an e-mail that it was "completely erroneous" to conclude that the relationship between the agency and its consultant had deteriorated. A spokesman for URS declined to comment.
Analysis 'very well done'
MnDOT has said it based its decision to postpone the replating because of URS' concerns that the work could actually weaken the bridge. According to the documents, however, URS did not raise this concern when it recommended the replating in the fall of 2006. Nor do the documents indicate that URS was focused on the integrity of the bridge's gusset plates.
Federal investigators have a "working theory" that the bridge collapsed in part because of the failure of gusset plates, which connected the bridge's girders.
"It is true that we don't have much safety margin for local stresses but on the other hand the overall stability of the structural system is not easily to be jeopardized," URS wrote as the company finished a draft report a year before the collapse. In handwritten notes from a September 2005 consultants' meeting, a URS official wrote that "gusset plate buckling -- if this occurs, it is not catastrophic."
From the start, URS executives were eager to do an exemplary job on their contract with MnDOT to evaluate the 35W bridge.
"I view this project as an important opportunity for us to get more work of this type from MnDOT in the future," Ed Zhou, a senior URS structural engineer, wrote in 2003. "This bridge will need deck replacement and rehabilitation."
For more than two years, URS consistently said that the bridge needed a major steel replating, according to the documents. That changed in a matter of weeks late last fall.
In the summer of 2006, URS formally recommended the replating project in its draft final report on the 35W bridge. MnDOT bridge design engineer Kevin Western told the consultant in an e-mail that "in general, the report and analysis was very well done."
At the same time, MnDOT raised concerns about the report's main recommendation.
"It's likely that retrofit of main members may be postponed for 15-20 years once an overlay is placed next year," MnDOT assistant bridge engineer Gary Peterson wrote in commenting on the draft report.
Instead of a major replating project, Peterson added, maybe MnDOT could simply "purchase plates and bolts for one or two member repairs ... should a short unrepairable crack be discovered."
Peterson added that in-depth weld inspections would continue every five years on critical members, and a "visual, arms-length" inspection would occur every year.
In earlier interviews, state bridge engineer Dan Dorgan said the replating project lost steam only following a January 2007 conference call when URS acknowledged that drilling required for the retrofit could in fact weaken the bridge.
More recently, Kender, MnDOT's spokesperson, said there were more factors in the agency's decision, with URS telling state officials that if inspections and testing showed no measurable cracks then their confidence in the bridge's structural integrity would increase.
When Peterson appeared to reject the replating idea, Flemming for one was taken aback. "Gary's comments seem more difficult to handle," Flemming wrote in an August 2006 e-mail.
"If we really experienced major crack propagation in a critical member that it may be too late to make such a repair in a timely manner," Flemming wrote, and then added that "I am very concerned that they are suggesting relaxing inspections to a five year cycle and that retrofit be postponed for 15 to 20 years."
Although Dorgan at the time stated that he remained committed to the replating, other comments by Peterson drew a concerned response from Zhou. Asked by Peterson at one point asked whether the bridge's "absence of crack[s] ... indicate a low probability of future cracking?" Zhou replied: "No one can explain why some fractures [on other bridges] happened after so many years," but added that "as we learned from probability and statistics, a zero-probability event may still happen."
As URS worked to finalize its report last November, Zhou was still committed to the retrofit plan. But e-mail exchanges show that by mid-December -- without explanation -- he had changed. "Based on all the results we have obtained, I strongly believe that doing a 2-million-dollar plating retrofit is not necessary," he wrote to Flemming on Dec. 13. Zhou did not return a phone call for comment.
Zhou instead recommended a plan to use electronic sensors to detect cracks. "After that, we should have the peace of mind just as good as doing the plating," he wrote. The idea was later rejected by MnDOT, which determined the technology was unproven.
In the end, URS recommended three options: the original replating plan; a plan to do some replating combined with inspections, and a plan to inspect the bridge and repair any cracks that were found. MnDOT chose the third option, which was described by URS as the "most cost efficient." And although they signed a contract to use URS in a May inspection of the bridge, state officials went ahead without the consultant.
Thirteen days before the collapse, MnDOT official Todd Niemann sent an e-mail to colleagues to tell them a meeting would be held in August "to determine if inspection can safely and adequately identify potential critical defects."
URS engineers continued to tinker with a final report on the bridge through e-mail exchanges on Aug. 1 -- just three hours before the 35W bridge collapsed.