U.S.
Baltimore bridge collapse highlights outdated safety standards around the nation, experts say
By
The Washington Post April 18, 2024
Michael Laris, Dan Keating and Júlia Ledur
The Washington Post April 18, 2024
Maryland Secretary of Transportation Paul J. Wiedefeld holds a briefing after the Key Bridge collapse. (Jonathan Newton/for The Washington Post)
U.S. standards for keeping bridges from collapsing when hit by ships hail from a different era.
They rely on half-century-old West German experiments on model ships for a key mathematical formula. Their minimum specifications cite the danger of empty 195-foot barges breaking loose from their moorings and drifting into bridges, a threat that seems quaint compared with the hulking 985-foot container ship that strayed off course after an electrical failure and toppled the Francis Scott Key Bridge in Baltimore last month.
And in 2020, international researchers warned that the standards sharply underestimate the impact of a head-on collision by a big ship into a bridge.
As federal investigators probe what caused the Baltimore collapse that killed six workers and shut down a vital U.S. port, some experts say the tragedy is shining a light on the need to bring bridge safety requirements into the modern era.
“The design standards take a while to update,” said Mark Gaines, who oversees bridges as Washington state’s design engineer. “It’s unfortunate, but sometimes it takes something like this as a wake-up call to say, ‘Wow, is this really the right standard?’ ”
There was nothing inherently wrong with the West German research, which was published in 1976, said Preben Terndrup Pedersen, a bridge strike expert and emeritus professor at the Technical University of Denmark. That early research had been focused on protecting reactors on nuclear-powered ships, and forms the basis for calculating the “ship collision impact force” on bridge piers in the U.S. specifications.
“But the numerical tools we have developed since then and the experience we have today are so much better,” Pedersen said. In a 2020 study in the journal Marine Structures, he and his colleagues found that U.S. specifications undercount the force of a collision with a large container ship by 40 percent and require far less resistance to ship strikes than the Eurocodes design standards used in the European Union. The study found the specifications were reasonable for calculating the force from some smaller ships.
The U.S. bridge design specifications are developed and issued by the American Association of State Highway and Transportation Officials, or AASHTO. Federal regulation requires that bridge designers use them on national highways. Officials at the U.S. Department of Transportation consult on the standards, but don’t have a vote on what they include, which is set by the association of state officials.
Patricia Bush, the association’s program manager for bridges and design, said group members she’s heard from had not seen the study. “It does look like that they did not do physical testing,” she said. She said her group’s state bridge experts will “certainly look at that one paper. But that doesn’t necessarily mean that their results are entirely accurate.” Pedersen said the team’s research included physical tests using models and was informed by real-world accident findings.
Maryland Transportation Secretary Paul J. Wiedefeld said in an interview he believes the state met the association’s standards — but they would have been the ones in place back in the late 1960s, when the bridge was being designed. “That’s a long time ago,” he said.
Back then, the minimum bridge design specifications did not include standards for how strong a bridge needed to be to withstand getting hit by a ship. Those didn’t come until 1994, according to the association, three years after its members drew up initial guidelines on the issue.
The Federal Highway Administration, the federal agency responsible for bridge safety, said in a statement that the “standards are routinely revised and updated for advances in knowledge, practices, and technology,” and that it is “confident in the reliability and level of safety of these standards.”
It also noted that while U.S. design specifications do identify an empty hopper barge as a minimum “design vessel,” the specifications also require that bridges in deep-draft navigable waterways be designed for collisions with the size and speeds of ships expected to travel under them.
It is possible to protect bridges from the megaships plying global trade routes. But doing so means broadening safety measures far beyond the bridges themselves and carefully coordinating different layers of safeguards, transportation experts said.
Earlier in his career, Pedersen joined fellow academics in opposing the plan for a massive Great Belt Bridge over the strait between the Danish islands of Zealand and Funen, a busy shipping channel. The risk of a bridge collapse was too great, he and others thought. He eventually came on board as a safety consultant, working with the government-owned bridge builder on designs for the soaring suspension bridge and a broader safety strategy for managing the risks from ships beneath it.
The goal was to keep the annual probability it would collapse at .0001, or once every 10,000 years.
To do that, they made the costly decision to move the piers much farther apart to provide a buffer between ships and critical bridge supports, and added sand islands to further protect the most exposed piers. They also established a “Vessel Traffic Service” to track ship movements and share information about weather and traffic conditions, and called for specialized pilots to guide the largest vessels past the bridge. They study near misses for safety lessons, and reconsider risks every 10 years to catch shipping and other changes, Pedersen said.
In the United States, with its vast geographic and political diversity, and layers of federal, state and local responsibility, a patchwork of oversight makes such coordination difficult. State transportation departments manage the structural safety of bridges, following the federally required guidelines for bridges on the National Highway System, while the Coast Guard is responsible for the safety of waterways and the risks related to shipping. States have their own bridge design standards, many heavily informed by the state highway and transportation officials association. A mishmash of port operators and international ship owners make things even more complex.
A few years after a cargo ship took down the Sunshine Skyway Bridge over Tampa Bay and killed 35 people in 1980, safety experts said the United States lacked a comprehensive approach to preventing such tragedies. The nation’s first design specifications addressing vessel strikes would eventually grow out of that Florida collision. But other problems identified following that incident have yet to be solved.
“Regulatory and institutional activities address parts of the ship-bridge-waterway system: none addresses the functioning of the system as a complex whole,” a National Research Council inquiry into vessel strikes concluded in 1983.
That has remained the case. And the lines of communication and responsibility were tested further after the Coast Guard was moved out of the U.S. Department of Transportation and into the new Department of Homeland Security, which opened in 2003.
States, meanwhile, often struggle to keep up with basic safety priorities.
In Washington state, for example, the transportation department doesn’t track marine traffic under its bridges or how it has changed over time, Gaines said.
Officials are busy with everyday threats — like holes opening up in a bridge deck — and securing the funding needed to address them.
“Vessel strikes like this are tragedies,” Gaines said. “But the biggest risk is just the basic maintenance and preservation of our network.”
In the frantic early hours after the Key Bridge crumpled into the Patapsco River, questions turned to whether Maryland had enough protection in place.
Wiedefeld, who heads Maryland’s transportation department, said the Key Bridge had some protective barriers, but officials were still looking into exactly what forces they were meant to protect against.
According to National Transportation Safety Board Chair Jennifer Homendy, the four barriers in the vicinity of the pier hit by the Dali are 25 feet in diameter. NTSB investigators have the original designs from construction in the 1970s. They are probing how barriers built according to today’s standards might perform differently, Homendy said.
Wiedefeld said he can’t opine on the ongoing NTSB investigation, which can take one to two years. But asked whether barriers could realistically have prevented the 95,000-gross-ton Dali from hitting and destroying the bridge, he suggested there are limits.
“The laws of physics are the laws of physics, right?” Wiedefeld said.
Federal transportation officials collect data from state inspectors on pier protections around bridges. But the data has focused on the presence — not necessarily the effectiveness — of such barriers.
For the Key Bridge, the pier protections were coded as “in place; functioning,” though they proved useless in preventing the span’s destruction.
Some places have taken other steps to compensate for the vulnerability of their bridges. In New York Harbor, large container ships are guided by tugboats operated by trained captains, said Vincent Barone, press secretary for the New York City Department of Transportation.
Ships heading in and out of the Port of Baltimore guide themselves under the Key Bridge with special pilots on board.
The Federal Highway Administration finalized new data specifications in 2022 for states to use, including a requirement for finer details on pier protections, which are set to take effect in 2026.
The presence or absence of pier protections does not provide a clear indication of whether a bridge can withstand a vessel strike. It’s here that the “Vessel Collision” provision of the latest edition of the bridge design specifications comes into play.
At its core is a formula for the “Annual Frequency of Collapse” — how often part of the bridge might be knocked down over the course of a year by a ship. The specifications say “critical or essential” bridges should withstand such dangers to a standard of one collapse per 10,000 years, the same as with the Danish span. The standard for “typical” bridges is one in 1,000 years. The Federal Highway Administration said it could not provide a list of bridges that fall into those “critical” or “typical” categories because the agency doesn’t collect that data; state agencies do.
To calculate the frequency of collapse, the specifications call for using tangible inputs like ship traffic under the bridge, but designers should also use the “probability of vessel aberrancy.” That is a measure of the risk the ship will face problems caused by pilot mistakes, poor environmental conditions or mechanical malfunctions.
It takes a vote by 35 of the 52 member agencies on the state highway and transportation officials association’s bridge committee to approve a change in the specifications, said Bush, the group’s bridge coordinator, noting that the votes rarely have more than a handful of dissenters and are frequently unanimous.
“NTSB will do a thorough investigation, like they do for all the bridge collapses. And if they come out with recommendations for AASHTO to change the specs, then we’ll take a look at those,” Bush said.
Jason Hastings, an engineer in charge of bridges at the Delaware Department of Transportation who sits on the association’s bridge committee, said there is “good reason” to move cautiously on changing the standards. Without establishing a clear scientific basis, Hastings said, being overly conservative about the size of ships that bridges should be designed to withstand could waste scarce resources.
“I know that right now there are 51 other chiefs of bridges and structures around the country that are looking at their bridge inventories,” Hastings said, considering “vessel collision along with all the other things that we have to be concerned with.”
He added: “We err on the side of caution. We err on the side of safety. So we’re not going to leave bridges open that are unsafe.”
Other states are watching closely. Louisiana “will follow the national lead once the investigation into the Baltimore incident is complete,” said state transportation department spokesperson Deidra Druilhet. The Federal Highway Administration said if the NTSB finds standards need to be revised, it “will work with the relevant AASHTO committees to develop and implement those revisions in a timely a manner.”
Researchers say changes are needed.
“The whole system was calibrated to older bridges, older ports, older ships, and so on,” said Roberto Leon, a structural engineering professor at Virginia Tech. “We have to recalibrate.”
He praised the association’s overall bridge expertise, but expects the Key Bridge disaster will spur improvement. “I think in probably two to three years, we’re going to see a complete revamp of this specification to bring it up to date to what we know today,” he said.