‘Nobody looks good’: Many entities failed before the Fern Hollow Bridge did

A nearly two-year federal investigation into the 2022 collapse of Fern Hollow Bridge in Pittsburgh found problems with every entity involved, from the city’s maintenance and the state’s oversight of the bridge to the private inspectors who were hired to keep tabs on its condition before the deteriorating southwest leg buckled, bringing the rest of the bridge down on a cold January morning.

That’s a summation of the numerous findings and analysis contained in more than 5,000 pages among 58 distinct investigative documents released last month by the National Transportation Safety Board, the federal agency investigating the bridge’s collapse.

Those documents include internal city emails and memos about bridge maintenance in the years before its collapse, expert analysis of the bridge’s crushed and crumpled steel, reviews of past inspections and load rating, and others.

They were all released in anticipation of the NTSB’s virtual board meeting set for Wednesday when it will meet to “determine the probable cause” of the bridge’s collapse, the NTSB said in a news release.

As the NTSB looks for who to blame, though, it will have to cast a wide net.

“There are no heroes in this,” Roberto Leon, a construction engineering professor at Virginia Tech University who has followed the bridge investigation and read through the NTSB documents, told the Pittsburgh Union Progress. “Nobody looks good. It’s a systemic failure.”

The documents show there was no single cause or entity responsible for the bridge’s collapse but rather a long series of missteps, mistakes and inaction by a variety of people and entities over decades that caused Fern Hollow Bridge’s failure.

“This is a good example of what happens when a number of things seem to conspire to go wrong,” said Kent Harries, professor of structural engineering and mechanics at the University of Pittsburgh, who has followed the bridge investigation and reviewed the NTSB documents. “It points out that bridges are a human endeavor that are built by humans, maintained by humans, imperiled by humans, and all the problems that go with that.” 

Some of the new findings spread among the 58 documents are stunning:

• When NTSB investigators began looking at the asphalt surface of the collapsed bridge, they found that the asphalt was about 6 inches thick. That is double the 3 inches that the original bridge design called for and double what inspectors assumed in their inspections and load capacity calculations previously used to set a 26-ton weight limit for Fern Hollow. That extra weight of more than 800,000 pounds could have at least resulted in a lower weight limit on the bridge if known before its collapse.
• When the steel cross-frame bracing under the western end of the bridge was deemed too corroded to help the bridge’s stability in heavy winds in 2008, the city infamously decided to install 1-inch-thick steel cables underneath both ends of the bridge to compensate. But an engineer with the firm that designed the cable system said it was never intended to be permanent, and the city was supposed to replace cross-frame bracing eventually. It never did.
• Part of the broader project the cables were part of was originally supposed to include painting parts of the bridge’s steel superstructure with a protective coating. But a city document shows that because it was nearing winter by the time that proposal was made in 2008, the city determined it would be too cold to paint and decided to do it later. But, as with the replacement for the cables, it never did.
• Investigators found that it appeared inspectors rarely, if ever, scraped off the corrosion from the sections of steel when they were doing inspections so that they could properly measure the thickness of the sections of deteriorating steel that were thinning because of the corrosion. Knowing the proper thickness of the deteriorating sections could have affected the bridge’s load analysis at a minimum.
• The truck that inspectors used to allow them to hang from a bucket while annually inspecting the underside of Fern Hollow Bridge was at least six tons heavier than the posted 26-ton limit on the bridge. But Tim Pintar, an inspector with one of the firms that inspected the bridge since 2005, said the firm never got a permit or approval from the city to use such a heavy truck there because the state said it was OK.

Finding the cause

As inspectors began looking over the crumpled heap of steel and concrete after Fern Hollow’s collapse, they very quickly focused on the southwest leg of the bridge, known by inspectors as “Bent 1 right leg,” or simply B1R. “Bent 1” refers to the two support legs and their cross-bracing under the western side of the bridge. Bent 2 is under the eastern side.

Outside experts had told reporters in the days after the collapse that the overhead pictures seemed to show that the bridge collapsed at the west end first and the southwest side was the starting point.

Part of that was confirmed early in February 2022 when NTSB released stills from cameras on the Pittsburgh Regional Transit bus that was on the bridge when catastrophe hit. The cameras showed the west end collapsing first.

Four months later, outside experts became even more convinced about the role the southwest leg played when copies of the bridge’s inspection reports dating back to 2005 included close-up photos that documented how the bridge’s frame legs and bracing were so badly corroded and deteriorating.

Those inspection report photos showed that the deterioration was significantly worse on the southwest leg, with a foot-wide hole in a web plate near the bottom of the leg, and significant thinning and deterioration in the last horizontal plate on that leg, called a “shoe tension tie plate” by inspectors.

NTSB and Federal Highway Administration investigators quickly focused on the same leg once they got to see all four legs when the collapsed bridge deck was removed.

“The distress observed in the Bent 1 frame legs, in addition to the commuter bus video, led FHWA investigators to consider the Bent 1, right, shoe as being principally involved in the collapse sequence,” an FHWA analysis examining Fern Hollow’s past bridge inspections and load ratings concluded in its report.

The “shoe” is the last section at the bottom of the steel frame legs that begins with the shoe tension tie plate. The shoe is attached to the concrete thrust blocks that were burrowed into the hillsides on either side of the bridge.

What inspectors found looking at the buckled southwest leg was that, compared to the three other legs, the outside portion of steel webbing, the vertical stiffener and the outside of the shoe tension tie plate were completely gone. 

While the other three legs all had corrosion in the shoe area of the legs, none of them was missing an entire section of steel web, vertical stiffener and shoe tension tie plate.

But each of the frame legs is “nonredundant,” meaning if one of them fails — like the southwest leg did — it would bring the rest of the bridge down with it onto Fern Hollow creek below.

“As we always thought, it certainly looks like it was” the southwest leg, said David Beck, a structural engineer in New Hampshire with more than a half-century of experience in design and construction.

And although it was not the overall cause, the 22-ton Pittsburgh Regional Transit bus that passed over the southwest leg right before the bridge collapsed that morning — “That was the trigger of the whole collapse,” he said.

Decades of problems

When the bridge’s inspection reports going back to 2005 were released to reporters a few months after the collapse, it was obvious from the year-by-year inspections why the bridge had deteriorated so badly: Small maintenance items inspectors flagged for the city to do, some as far back as 1995, were almost never done.

“And the No. 1 problem was the clog scuppers and downspouts on almost all their bridges. And I tried to preach that that whole time, and nothing ever got done. I mean, just simply clean the stuff, you know?” Tim Pintar, a CDM Smith bridge inspector who had inspected Fern Hollow five times, told NTSB and FHWA investigators in an interview in August 2022. A transcript of the interview was in the NTSB documents released last month.

Those maintenance items, in one way or the other, related to trying to keep water — and wintertime salt-laden water in particular — from leaking off the bridge deck onto the steel superstructure below.

The reason that was so critical with Fern Hollow Bridge is that designers in 1973, when the bridge was built, chose to use what is known as “uncoated weathering steel” for the bridge’s superstructure.

Because this kind of steel goes unpainted — leaving behind a planned patina of brownish rust that should protect the steel if properly maintained — it leaves it at risk of corrosion and deterioration if the steel continually gets wet, as happened with Fern Hollow because of drainage issues.

The maintenance issues, as Pintar noted, included cleaning the bridge scuppers, or drains, regularly rather than letting the scuppers get full of debris, pooling water on the bridge, which then ran off its sides and down below through cracks in the bridge deck’s concrete.

Other maintenance items that were regularly recommended but not done were similar: seal the bridge deck so water cannot leak through to the superstructure below; extend the weep holes, or drains below the bridge, so the water would drip away from steel; paint the frame legs and steel joints with a special coating to protect the steel from falling water. 

But a question NTSB and FHWA inspectors dug into over the past two years was this: If the city continually ignored maintenance recommendations, possibly because of the city’s budget constraints, why didn’t inspectors simply raise alarm bells by making those basic recommendations either “critical” or a “high priority”?

Under the state’s bridge inspection system, when inspectors find problems that the bridge owner needs to fix, they can give the problem, such as clogged scuppers or painting frame legs, a ranking from 0 to 5.

Ranking a problem 0 — “critical” — or a 1 — “high priority” — puts those problems not only on a clock to be fixed — they have to fix a 0 ranking within seven days and a 1 ranking within six months — but also, maybe more importantly, it brings state inspectors in to make sure those problems are fixed quickly.

Priority 2 items do have to be completed before the next regular inspection. But problems ranked from 2 — “priority” — to 5 — “routine” — do not bring state inspectors to town to make sure they get done. 

Essentially the ranking system gives inspectors the power to get a local bridge owner to act, whether they want to or not. Federal investigators seemed perplexed why inspectors or the state’s own inspectors did not use that power.

In the interview of Pintar, investigators discussed how inspectors identified the significant and growing deterioration of the cross-frame bracing on the western end of the bridge — Bent 1.

Steve Prouty, senior structural engineer for the National Transportation Safety Board, asked Pintar: “So, and this is just kind of thinking off the cuff, here, so if nothing was getting done with it, did it cross your mind to maybe, you know, elevate the priority of that?”

Pintar responded: “We sent — like I said, we had zeros and ones; and then, they did finally remove the bracing at the bottom because they were going to fall down onto the trail.”

But while that cross-frame bracing was eventually removed in 2019 after it became completely severed from the frame legs, it was only removed after inspectors ranked it 0, or a “critical” problem that had to be repaired within seven days.

A similar ranking was never assigned, however, to repairing the ever-growing deterioration of the lower frame legs, problems that were never ranked higher than a 2, which meant the state was never brought in to ensure they were repaired.

Load rating and fracture problems

Investigators were similarly confused why inspectors, as well as state inspectors who review bridge inspections when they are completed, never requested a new load rating analysis in the eight years before the bridge’s collapse as corrosion and deterioration got worse and worse.

A new load rating analysis could have lowered the bridge’s 26-ton weight limit set in 2014 or even forced the bridge’s closure.

Justin Ocel, an FHWA senior structural engineer, asked Pintar: “But as the inspector, you do make recommendations to the owner, so why not recommend a load rating if you’ve seen a change of condition?”

Pintar, who worked on Fern Hollow inspections in 2007, 2009, 2014 and 2015, as well as a fatigue and fracture plan for the bridge in 2016, replied: “Probably, because they felt like it wasn’t enough of a change in condition to affect it.”

In a separate interview by federal investigators last year, Rich Runyen, assistant chief bridge engineer for the Pennsylvania Department of Transportation, which oversees all bridge inspections in the state, said even if inspectors did not request a new load rating analysis, the state itself could order one after reviewing an inspection report.

In fact, he said, after looking over Fern Hollow’s inspections from the years prior to its collapse, he was surprised no one, from the inspectors to the city or state, ever requested a new load rating analysis after 2014.

But Runyen also said that that while PennDOT’s inspectors do regularly read the inspections of state-owned bridges “cover to cover,” he added: “I cannot say with confidence that we are doing cover to cover on every single inspection” for locally owned bridges, such as Fern Hollow, which is owned by the city of Pittsburgh.

Federal investigators also found that there were multiple specific mistakes by inspectors that contributed not only to a failure to request a new load rating analysis but also to lower the bridge’s overall rating.

One of those failures was that inspectors over multiple years seemed to ignore the importance of the shoe tension tie plate — the last horizontal plate that marks the top of the shoe at the bottom of the frame legs.

Investigators showed through an analysis of the engineering forces that the shoe tension tie plate played a unique role in the bridge’s structure. Not only was the shoe tension tie plate thicker than the horizontal tie plates above it on the frame legs, but also it was specifically designed to handle more of the forces from the bridge above.

But as photos in the various inspections showed, while inspectors focused on the growing corrosion and hole that was forming in the web plate beneath the tension tie plate, they did not seem to appreciate the significance of the corrosion that was slowly thinning the thickness of the shoe tension tie plate nearby at the same time.

NTSB mapped the corrosion on the southwest leg’s shoe tension tie plate and found that it was less than half the thickness it should have been by the time the bridge collapsed.

Given all of that, the NTSB concluded that “if the tie plate had been properly accounted for in the load rating, it would have controlled the rating and possibly resulted in closure of the bridge.”

That tie plate “was a significant part” of the collapse, Harries of Pitt said. 

But that revelation also shows that a lack of historical knowledge may have played a role in the bridge’s failure, he said.

“The tie plate is one example that I’m not certain whether the inspectors or the people reviewing the inspections, or doing the load ratings, had sufficient information about how this bridge worked” in its engineering forces, he said. “The designer knew how it worked [when it was built in 1973], but when it was handed off two, three or four generations later, how was that institutional knowledge passed on?”

Another failure in the bridge’s inspection history was that after 2011 — for reasons that are not clear to investigators — Fern Hollow’s four frame legs were not considered to be “fracture critical members” of the bridge’s superstructure, even though they were “nonredundant.” A fracture critical classification would have meant more focus on the frame legs than they were given, possibly resulting in a new load rating, lower bridge rating or even closure.

Pintar, who was the lead engineer on the bridge’s 2016 fatigue and fracture plan, like other inspectors before him, only identified the bridge’s steel girders and floorbeams that supported the bridge deck as fracture critical.

Prouty asked him: “So, did you think it was odd that this bridge didn’t have — you know, the legs weren’t considered fracture critical members?”

Pintar answered: “I still wrestle with that” before explaining that, as he understood it, the frame legs were considered only to be “in compression,” which prevented them from being labeled fracture critical because they were not “in tension.”

Being “in compression” would mean the frame legs were being squeezed by forces from above and not “in tension,” meaning they stretched to absorb those forces. 

But the FHWA’s analysis found the frame legs had “elements” that were “in tension” and should have been considered fracture critical. 

The NTSB factual report concluded: “Since the frame legs had elements in tension and were nonredundant, the frame legs should have been considered fracture critical members in the Fern Hollow Bridge inspection reports but were not.”

If the frame legs had been deemed fracture critical, the state’s bridge manual would have dictated that any holes due to corrosion in a fracture critical member be ranked a 0, or “critical” problem that had to be fixed within seven days with state inspectors overseeing the repairs.

The fact that Pintar and other engineers from the city, the state and the two engineering firms who inspected the bridge since 2011 all got it wrong that the frame legs were fracture critical was stunning to Leon, the Virginia Tech professor.

“At some point there somebody should have stopped and said: ‘What’s happening here?’” Leon said.

He noted that not only would an inspector like Pintar, who is an engineer, have reviewed the bridge’s forces to see if the legs were fracture critical but also so would another engineer either with every inspector’s firm, or the state.

“It’s supposed to be a complex system with checks and balances,” Leon said. “In this case, our checks and balances failed.”

RELATED STORY: State, local and federal officials have changed bridge care since Fern Hollow collapse