[Note that this article is a transcript of the video embedded above.]
On March 26, 2024, the Francis Scott Key Bridge in Baltimore, Maryland, suffered a catastrophic collapse after being struck by a large container ship. This event, just weeks prior to this writing, resulted in the tragic loss of six construction workers’ lives, injuries to another, and major disruptions to both road and sea traffic. The news coverage was extensive, and many independent creators have provided insightful analysis. While investigations are ongoing, from an engineering perspective, it’s crucial to understand how vessel collisions are considered in bridge design and why the Francis Scott Key Bridge succumbed to this incident. I’m Grady, and this is Practical Engineering. Today, we’re examining vessel collision design for bridges and drawing parallels to unexpected failures we sometimes see even in seemingly robust systems – much like diagnosing complex issues in a project car or even the surprising complexities hidden within seemingly simple “dixie cup coding” solutions.
The Francis Scott Key Bridge was a continuous through truss bridge, a complex structure combining different engineering principles. To understand this better, let’s turn to our correspondent, Road Guy Rob, for a breakdown of bridge terminology.
Well, Grady, I’m in Long Beach, California, at the site of the new bridge replacing the old Gerald Desmond Bridge, which was a similar arch/truss structure, almost a smaller version of the Key Bridge. The Port of Long Beach is relieved to see the old one gone.
The Gerald Desmond Bridge was a truss bridge. Unlike a solid beam, a truss uses many smaller interconnected members to create a rigid and lighter structure. This efficiency is a hallmark of clever engineering. Both bridges were “through-truss” designs, meaning the roadway is suspended beneath the arching truss. The Key Bridge was a continuous truss, a single, rigid piece spanning its entire length, distributing load across all three spans.
The approach roads are separate bridge structures, simple girder spans supporting concrete roadways.
The old Gerald Desmond Bridge in Long Beach? It’s been demolished, its remnants consigned to history. In its place stands the new, cable-stayed Long Beach International Gateway. The improvements implemented in this new bridge project offer valuable insights for Baltimore as they plan the Key Bridge replacement.
More on that shortly.
When the container ship Dali lost power and collided with the southwest pier of the Key Bridge, the support failed, causing the southwest and central spans to collapse vertically. A section of the northwest truss detached and rotated, pulling several approach spans with it. Fortunately, a mayday call from the ship allowed police to halt traffic, but tragically, eight construction workers on the bridge could not be evacuated in time, resulting in six fatalities.
The salvage operation is immense, involving multiple floating cranes, including the renowned Weeks 533. It’s like a giant Jenga game; each cut and lift requires precise planning and execution due to the underwater debris and unpredictable stresses in the wreckage. Baltimore’s port facilities are crucial for processing the removed materials. Reopening the port is paramount. A limited channel has been opened, and the Army Corps of Engineers is working to restore full access, but the timeline remains uncertain. Road traffic is also heavily impacted, with detours causing significant delays, especially for hazardous materials transport restricted from the harbor tunnel.
While a full NTSB report may take a year or more, the failure mode seems evident. A massive ship hitting a bridge pier caused the collapse. The less obvious aspect is how engineers account for such events in design to prevent them. Vessel collisions are a known risk. “Allision,” the term for a moving vessel striking a stationary object, is common in maritime and bridge engineering. From 1960 to 2014, 35 major bridge collapses were attributed to vessel impacts, 18 in the US, highlighting the risk in areas with extensive waterways.
The 2001 Queen Isabella Causeway collapse in Texas and the 2002 I-40 bridge collapse in Oklahoma, both caused by barge collisions, and the 2009 Popp’s Ferry Bridge collapse in Mississippi, also from barges, underscore this vulnerability. The 2012 Eggner’s Ferry Bridge collapse in Kentucky from a cargo ship further emphasizes the point. However, the 1980 Sunshine Skyway Bridge disaster in Florida, where a ship collision caused the main span to collapse and 35 deaths, significantly heightened awareness and spurred critical changes in bridge design standards. But horizontal collisions aren’t the only threat. Rob explains vertical clearance concerns.
The Key Bridge failure was due to a horizontal allision, a side-to-side impact.
In Long Beach, vertical clearance was the primary concern for the old Gerald Desmond Bridge. At 155ft tall, it seemed sufficient in 1968, but modern ships are much larger. The new bridge is over 200ft tall, a 50ft increase to accommodate these larger vessels.
The Long Beach International Gateway eliminated the port’s undesirable distinction of having the shortest bridge clearance in the US. This upgrade was proactive, unlike the tragic circumstances in Baltimore.
Following the Sunshine Skyway collapse, federal agencies and engineering bodies from both maritime and bridge sectors collaborated to update bridge codes, incorporating vessel collision considerations. In the US, AASHTO sets these standards for highway bridges, with similar regulations worldwide, including the Eurocode.
Infrastructure design often involves “worst-case scenarios,” but practicality sets limits. Designing for every imaginable extreme, like meteorite impacts, is infeasible and cost-prohibitive. The same applies to ship collisions. The sheer kinetic energy of modern cargo ships is immense. Building bridges to withstand absolute worst-case collisions would be financially crippling. Instead, for “high consequence, low probability” events, codes define acceptable risk levels. There’s always some risk, and society must balance safety with the benefits of navigable waterways. For critical structures like the Key Bridge, AASHTO sets a 0.01% annual probability of collapse from vessel impact – roughly the annual chance of rolling a Yahtzee. This means engineers design for a low probability of collapse, not guaranteed survival in every scenario.
The annual probability of bridge collapse from ship collision is calculated using a formula incorporating several factors: the number of ships passing annually, a growth factor for future traffic, the probability of vessel aberrancy (loss of control), and the geometric probability of collision if a ship veers off course. The geometric probability depends on the distance of piers from the navigation channel. Wider separation reduces risk. Tug escorts can also mitigate aberrancy probability, a measure not required at the Key Bridge.
Even with a collision, collapse isn’t inevitable. Structural engineering plays a vital role. Collapse probability depends on pier strength and impact force. Impact force isn’t a static load; it varies with vessel size, speed, ballast, and collision angle. It’s typically simplified to an equivalent static load, calculated using vessel deadweight tonnage and velocity. The formula, while simplified, conservatively assumes fully loaded vessels.
For example, the Sunshine Skyway Bridge ship had a 34,000-tonne deadweight. At an estimated 5-knot impact speed, the force was roughly 56 meganewtons (13 million pounds). The Dali, with 117,000 tonnes deadweight and 5-knot impact speed, generated over 100 meganewtons (24 million pounds), assuming full load. Back-of-the-envelope physics confirms these magnitudes. Decelerating the Dali from 5 knots to a stop in 4 seconds requires approximately 72 meganewtons (16 million pounds). This is equivalent to the force of five SpaceX Starships hitting a pier – a truly immense force.
Designing piers to withstand such forces is complex and costly, affecting the entire load path, including foundations and the superstructure. Besides pier strength, engineers employ pier protection systems. Fenders cushion impacts, but for large ships, they’d be enormous. Islands force grounding before pier impact, but create environmental issues and wider spans. Dolphins, circular sheet pile structures filled with material, deflect or stop ships. The new Sunshine Skyway Bridge uses islands and dolphins. The Key Bridge had dolphins, but the Dali’s angled approach bypassed them.
It’s crucial to remember that these modern vessel impact considerations were absent when the Francis Scott Key Bridge was built in 1977. No major upgrades triggered adherence to newer codes. Ships like the Dali were nonexistent until around 2006. Future ship designs are also unknown. Hindsight suggests better protection, but applying this retroactively to all similar bridges represents a massive resource investment for statistically rare events. This isn’t to say it’s unwarranted, but such decisions are complex. Code changes might not be immediate, but vessel collisions will be central to the Key Bridge replacement design. Rob elaborates.
Visually, the Key Bridge appeared well-maintained. If the NTSB report confirms good bridge condition and attributes the collapse solely to the ship, it wouldn’t be surprising.
The old Gerald Desmond Bridge in Long Beach, however, had structural issues. Its environmental impact report revealed significant structural deficiencies. Bridges are scored out of 100; a new bridge gets 100. The old Gerald Desmond scored 43. Scores below 80 qualify for federal rehabilitation funds; below 50 triggers replacement funding.
Salt air corrosion was a major factor, degrading the paint protecting the steel members. Paint failure is critical in truss bridges, where every member is vital.
Load analysis showed main span members overstressed even by standard trucks. Concrete cracking was critical, requiring nets to catch falling debris.
Long Beach had four objectives for the new bridge, relevant to Baltimore’s situation: 1) 100-year design life. 2) Reduced approach grades – achieved with long viaducts. Baltimore’s approaches are already long, so changes are less likely. 3) Increased roadway capacity – six lanes versus four on the old bridge and the Key Bridge. The new bridge also has shoulders, improving safety. 4) Vertical clearance for larger vessels. Baltimore might reframe this as horizontal clearance, prioritizing allision prevention.
Thanks, Rob. Check out his channel for more on transportation infrastructure. This tragedy wasn’t solely a bridge failure; it stemmed from a maritime navigation failure. Bridge engineers have limited control over vessel safety and navigation. To prevent recurrences, a holistic approach is needed, addressing both structural solutions and maritime risk reduction. This might include tug escorts for large ships near vulnerable bridges. The NTSB report should offer recommendations.
Risk analysis, while essential in engineering design, can be unsatisfying. Humans struggle with probabilities, and communicating risk effectively is challenging, leading to terms like the “hundred-year flood” being misunderstood. It’s unsettling to acknowledge that even code-compliant bridges carry inherent risks. Everything involves tradeoffs, but the public affected by these risks often has limited input on acceptable risk levels. The question isn’t simply “Can we build safer bridges?” – the answer is always yes. The real questions are “How much risk is tolerable?” and “How much are we willing to pay for incremental safety improvements?” Requiring all bridges to withstand worst-case collisions would drastically reduce bridge availability. Events like this serve as stark reminders that risks aren’t abstract numbers; they have real-world consequences. My condolences to the victims’ families. Learning from this tragedy is crucial to improve both infrastructure and maritime safety, preventing future occurrences. This event highlights the complex interplay of engineering disciplines and risk management, reminding us that even seemingly robust structures, like bridges or even well-engineered project cars, are vulnerable to unforeseen events. And just as “dixie cup coding” can sometimes mask underlying complexities, simplified risk assessments can sometimes obscure the true magnitude of potential consequences.