Within six hours of the Florida International University pedestrian bridge collapsing, killing six, Gov. Rick Scott’s press office sent out a “fact sheet” that claimed his Department of Transportation (FDOT) had little to do with the project.
In emails obtained by the NBC 6 Investigators, members of the FIU’s project team almost immediately expressed “shock” at the governor’s “misstatements and misinformation,” calling the press release “fake news” and saying FDOT had “extensive involvement.”
“This is so not accurate and we were shocked” by FDOT’s claims, wrote Ken Jessell, FIU’s chief financial officer, to a senior staff member at the state university system.
“FDOT was involved every step of the way…Every plan was reviewed and signed off by FDOT,” he said.
In its March 15 statement, FDOT said it did just a “routine preliminary review” of the project, while issuing traffic permits, passing through federal money and authorizing FIU to occupy the space above Eighth Street for what was to be its showcase concrete pedestrian bridge.
But records obtained by the NBC 6 Investigators suggest the state was much more involved and concerned than it has let on.
FDOT engineer Thomas Andres, who oversees bridge design plans, informed FIU’s design-build team nearly two years before the bridge came crashing down that avoiding cracking in the bridge during construction would be difficult.
Commenting on the bridge’s preliminary plans on March 25, 2016, Andres wrote “maintaining stress limits throughout all intermittent phases to avoid cracking of the members will be extremely tricky” and he expressed concern that the early design may not have adequately compensated for “shear” – the very force that engineers say appears to have led to the bridge’s collapse.
What, if anything, FIU’s design-build team - FIGG Bridge Engineers and Munilla Construction Management (MCM) did in subsequent designs to address FDOT’s concerns is not yet known.
The companies say they are prohibited from commenting while they participate in the National Transportation Safety Board investigation into the failure.
FDOT has not responded for weeks to requests for emails and other records that may shed light on what FDOT did or did not do to make sure their concerns were addressed. Nor would FDOT or Andres answer questions about Andres’ role and how it may conflict with the state’s claim that FDOT was barely involved in the design plans.
An NBC 6 public records request with FIU for all emails mentioning Andres’ name and any responses to his concerns is also pending.
But this we did confirm: A telephone message from FIGG’s bridge engineer, Denney Pate, informing FDOT of cracking in the bridge two days before the collapse was left with Thomas Andres - the same engineer in the state structures design office who early on expressed concerns about potential cracking and design issues.
Pate described for Andres “some cracking that’s been observed on the north end of the span...and obviously some repairs or whatever will have to be done but from a safety perspective we don’t see that there’s any issue.”
FDOT said Andres did not listen to the voicemail until March 16, the day after the collapse, and FDOT and Andres have so far declined to say what Andres might have done had he heard it, including possibly ordering closure of Eighth Street.
The NBC 6 Investigators showed Andres’ comments, along with plans, calculations and other documents obtained through public records requests, to Al Brizuela, a Miami-based professional engineer.
“He’s thinking this is going to be problematic,” Brizuela said, referring to how the complex, aggressively designed truss bridge was going to be stressed before it was lifted, rotated and placed on supports above Eighth Street.
That operation was completed the morning of Saturday, March 10, and severe cracking was documented on the north end of the span within days.
The failure on March 15 came in that same area, as a work crew was carrying out previously unplanned re-stressing of steel bars that ran through a diagonal truss on that north end of the 175-foot span – an attempt, some engineers have theorized, to close the cracks that had formed on that truss.
The process is called “post-tensioning” and it is one of the things Andres warned about two years earlier, writing “predicting where the (post-tensioning) stressing actually goes will be tricky.”
While Brizuela cautions that not all the plans have been released – so any opinion is not yet fully informed - he questions whether the final plans and actual construction took into account Andres’ concerns about the careful sequencing of stressing bars and cables that run through the deck, canopy and trusses.
Even before the apparently fatal decision to re-stress the truss in the area of the failure on the day of the collapse, there is conflicting information in the available record about the sequence of stressing.
For instance, calculations done by FIGG dated Feb. 10, 2017, assumed stressing would begin with tendons in the canopy and post-tension bars in the two outermost diagonal trusses, followed by tendons in the deck.
But plans dated April 7, 2017 - marked “for construction” and signed and sealed by FIGG’s engineer Pate – call for the tendons in the deck to be stressed first, followed by the canopy and then the post-tension bars in the outermost trusses.
Also, a February 2018 memo from Bolton Perez & Associates, engineers hired by FIU to inspect the design-build team’s work, states the northernmost diagonal truss was stressed a day before its southern-end counterpart in late January, in preparation for the move to above Eighth Street. But the plans released so far call for those trusses to be stressed in the opposite order.
“They did it backwards,” Brizuela said, though again cautioning that is only an assumption based on the records available publicly so far. Another record appears to state they were stressed on the same day.
Moreover, it is not known whether altering those stressing sequences would make a significant difference in calculations conducted to determine if the structure can withstand the forces being applied to it during and after construction.
But, in his comments, Andres refers repeatedly to the importance of adopting the proper stressing sequence.
He said all 12 members of the main span “may have to be stressed to avoid cracking.” The “for construction” drawings released so far reveal that was not done for all 12 members.
Andres also warned that “the web truss will be very difficult to form without shrinkage cracking” on some of the members.
Cracking was indeed observed on two of those members on February 6 - one week after tension was applied to adjacent trusses, as the main span was being prepared on the roadside for lifting and movement to its resting place over the highway. Further, larger cracks were found weeks later on each end of the main span before the move.
Another concern of Andres that may not have been adequately addressed in subsequent plans: an 8-inch diameter drain pipe running down the bottom center of the deck “will likely create a weak point which will be a crack initiation point,” Andres noted.
Brizuela said the “for construction” plans he reviewed did not show any significant change in the design of the drain pipe, which remained at the same location and the same size.
Andres submitted his comments on the preliminary design into an FIU document called an “over-the-shoulder review” on March 25, 2016, according to records released by FIU.
He stated the comments were “for information only. No response is required. The comments are intended to assist in providing general feedback to the (design-build firm).”
Four weeks later, a FIGG bridge engineer acknowledged the comments, saying they were “intended to assist in progressing the (design-build firm’s) concept to 90% plans.”
On April 25, 2016 the record reflects Andres accepted that response and closed the comments section.
Taken as a whole, Brizuela said, Andres’ comments point to some of the same issues that he believes ultimately contributed to if not caused the bridge to collapse.
“He’s pointing out what’s going to happen. What could happen,” Brizuela said, concluding, “Unfortunately it looks like he was kind of predicting what was going to happen.”
Given Andres’ concerns about potential cracking and shear forces built into the original plans, Brizuela wonders what Andres would have done had he received Pate’s March 13 voicemail.
“I think he would’ve stopped traffic,” Brizuela suggested. “That’s the only thing you can do at that point. You want to protect the public.”
FDOT said it possesses no email discussing cracks in the week before the collapse, though FDOT’s local agency project assistant coordinator did attend a meeting held on the morning of the collapse where Pate and others, from MCM and FIU, discussed the cracks.
After that meeting adjourned, a crew from Structural Technologies was carrying out orders to re-stress bars in the northernmost diagonal truss, directly above where the most severe cracks had formed, when 950 tons of concrete collapsed onto the road below, killing a Structural employee who was working on the canopy and five people in cars below.
In addition to Andres not listening to his voicemail for three days while “out of the office on assignment,” there’s also no evidence he saw photographs of the cracks taken on March 13 by MCM and March 14 by Bolton Perez & Associates.
Brizuela said any engineer who did see those photos or the cracks themselves should have heard a clear message: “The bridge is failing. Once you see cracks (such as those in the photographs) in concrete, we know there’s a failure occurring within the concrete.”