How Bridge Engineers Design Against Ship Collisions

On March 26, 2024 (just a few weeks ago, if you're  watching this as it comes out), a large container  

ship struck one of the main support piers of  the Francis Scott Key Bridge in Baltimore,  

Maryland, collapsing the bridge, killing  six construction workers, injuring one more,  

and seriously disrupting both road and marine  traffic in the area. There’s a good chance you  

saw this in the news, and hopefully you’ve seen  some of the excellent content already released by  

independent creators providing additional context.  I got a lot of requests to talk about the event,  

and I usually prefer to wait to discuss events  like this until there are more details available  

from investigations, but I think it might  be helpful to provide some context from an  

engineering perspective about how we consider  vessel collisions in the design of bridges  

like this one, and why the Francis Scott  Key bridge may have collapsed. I’m Grady,  

and this is Practical Engineering. Today we’re  talking about vessel collision design for bridges.

The Francis Scott Key Bridge was a steel  arch-shaped continuous through truss bridge.  

I’m working on a video that goes into a lot more  detail about the different kinds of bridges and  

how they’re classified, but this bridge had kind  of a medley of structural styles, so let me hand  

it off to our special guest correspondent,  Road Guy Rob, to break that terminology down.

Well, Grady, I'm in Long Beach, California  today, standing on top of this brand new  

bridge that replaced an old arch/truss  bridge that used to be right there. It  

kind of looked like a baby Key Bridge, and the  Port of Long Beach is happy that it's gone.

The Gerald Desmond Bridge was a truss  bridge. Instead of having one big large beam,  

a truss has lots of smaller connected  structural members all attached together.

This creates a rigid structure that's  much lighter weight than a big heavy beam,  

and that makes trusses efficient and clever when  they work. Both the Key bridge and the old bridge  

that used to be here were “through-truss”  bridges. It's a sort of arch shape,  

and the driving deck is suspended below the  truss, so you sort of drive through the arch,  

but it's not an actual arch with like a keystone  and all the pieces pushing horizontally to hold  

each other together. No, this through-truss bridge  has no hinges or joints at the main supports,  

nothing that breaks it up into sections. So  that's why engineers called the Key bridge a  

continuous truss bridge. It's all one big piece,  and it's all bolted and welded into a single rigid  

truss across its entire length. And then that load  distributes across all three spans of the bridge.

Now, the approach roads on each side are  entirely separate bridges, even though they  

link together. They just look like concrete  roads sitting on top of simple girder spans.

Well, you ask, what happened to that baby Key  bridge in Long Beach? Well, the only way you're  

going to see it now is to turn on Grand Theft  Auto five and look at the fictionalized version  

of it immortalized in Los Santos for all time.  Because when this bridge opened, the port of Long  

Beach demolished the old bridge and the last  scraps of it got all the way back in October.

In its place, this new, fancier  looking cable-stayed bridge,  

the Long Beach International Gateway. And what  the Port of Long Beach did in studying to build  

this bridge and the list of improvements they  came up with might give us some clues what  

Baltimore might want to end up doing when  they replace the Key bridge down the line.

And we'll talk about that in just a moment.

When the Dali container ship lost power  and drifted into the southwest pier,  

the support collapsed, and most of the truss  and deck fell with it. Both the southwest and  

central spans fell roughly vertically with the  loss of support from the damaged pier. Part of  

the truss on the northwest side separated  from the unsupported section and rotated  

toward the northeast span, taking several  of the approach spans with it. Thankfully,  

the ship had put out a mayday call before  the impact, allowing police officers at  

either end of the bridge to close it to traffic.  Tragically, it wasn’t enough time to get the crew  

of eight construction workers off the structure  before it fell, six of whom lost their lives.

Just dealing with the salvage and removal of  the steel and concrete debris left over from the  

collapse has been a massive undertaking. Within  a week, engineering teams were on-site measuring,  

cutting, lifting, and floating away huge chunks  of the wreckage in separate salvage operations for  

the main bridge, approaches, and the vessel. As of  this writing, they’re still working hard on it. At  

least seven floating cranes were involved,  including the famous Weeks 533 that pulled  

US Airways Flight 1549 from the Hudson River in  2009. This was essentially a massive Jenga tower:  

the order of operations and the precision of  each cut and each lift mattered. With so much  

debris underwater, they had to map it out to  understand how everything was stacked together.  

Access was a major challenge, and the stresses  in the wreckage were hard to characterize,  

so it’s been a slow and deliberate process  requiring careful planning and tons of skill  

to do safely. Fortunately for Baltimore, there  are large industrial facilities in the port that  

can process the thousands of tons of material  that will ultimately be removed. Of course,  

reopening the port to shipping traffic is a huge  priority. A small channel was marked out under  

one of the approach bridges for smaller vessels  like tug and barges traffic, and the Army Corps  

of Engineers is making good progress on opening  up the main channel, but it isn’t clear when  

full-scale operations at the port will be able to  resume. Shipping traffic isn’t the only traffic  

affected either, the bridge carried thousands  of road vehicles per day that now have to be  

re-routed. There is a tunnel under the harbor that  provides a decent alternate route, but trucks with  

hazardous materials aren’t allowed through,  requiring an enormous detour around the city.

It’s been more than a month since the event,  but it will likely be a year or more before  

we get an official report documenting the  probable cause of the failure. In the US,  

events like this are investigated by  the National Transportation Safety  

Board or NTSB. This independent government  agency is extremely diligent. And often,  

diligent also means slow. But events like this  are how the field of engineering evolves. Human  

imagination isn’t limited to past experiences,  but in many senses, engineering is. We just don’t  

have the resources to answer the millions of  “what ifs” that might coalesce into a tragedy,  

so we lean on the hard lessons learned from  past failures.

When something terrible happens,

it’s really important that we collectively get  to the bottom of why and then make whatever  

changes are appropriate to our engineering  systems to prevent it from happening again.

But, at the risk of stating the  obvious, the failure mode in this  

case is pretty clear. You probably don’t need an  engineer to explain why a massive ship slamming  

into a bridge pier would cause that bridge to  collapse. I think what’s less obvious is how  

engineers characterize situations like this so  that bridges can be designed to withstand them.  

Collisions with bridges by barges and ships  are not a modern problem. Technically they’re  

called “allisions” since a bridge isn’t  moving, but that term is used more in the  

maritime industry than by bridge engineers.  Between 1960 and 2014, there were 35 major  

bridge collapses resulting from vessel impacts.  And, 18 of those were in the US. We just have  

such a big network of inland waterways,  and that means we have a lot of bridges.

Two spans of the Queen Isabella Causeway Bridge  in Texas collapsed in 2001 when barges collided  

with one of the piers. A year later, a bridge  carrying I-40 over the Arkansas River in Oklahoma  

was hit by barges when the captain lost control,  collapsing a major portion of the structure. In  

2009, Popp’s Ferry Bridge in Mississippi collapsed  after being struck by a group of barges. In 2012,  

the Eggner’s Ferry Bridge in Kentucky fell  when a cargo ship tried to go through the  

wrong channel. Before any of those, though,  the Sunshine Skyway Bridge in Florida put a  

major focus on the problem. In 1980, a bulk  carrier ship lost control because of a storm,  

crashing into one of the piers and collapsing the  entire main span of the southbound bridge, killing  

35. The event brought a lot of new awareness and  concern about the safety of bridges over navigable  

waterways. But piers aren’t the only parts of a  bridge at risk from ships. I’ll let Rob explain.

The Key bridge got into trouble because  of a horizontal allision. That's where a  

ship moves side to side in the wrong way  and hits something it's not supposed to.

Here in Long Beach, that really wasn't  their concern, primarily because the old  

bridge columns were way inland here, so there  was no way for a ship to exit the waterway  

and hit the column because the column was in  lots of dirt. And the new replacement bridge  

takes no chances at all. Look how much  farther onshore those columns are now!

Now, the Port of Long Beach were far  more worried of the old Gerald Desmond Bridge  

getting hit vertically. The old bridge was 155ft  tall. That's like a 15 story building. And if that  

sounds pretty tall to you, it sounded pretty tall  to them back in 1968 when they built the bridge.  

But as we now know, ships are getting bigger and  fatter and taller and 155ft wasn't cutting it for  

some of the modern ships that were trying to get  into the back part of the port, where there's a  

lot of cranes and action happening over there.  So the new bridge adds another 50ft, takes it  

over 200ft. That's like a 20 story building to  get up from the waterline to that new bridge.

And this new, taller, Long Beach International  Gateway helps the port scratch off one designation  

they didn't want - having the shortest bridge  over a port in the United States. Well,  

that's gone now, and thankfully in a less tragic  manner than what's happening on the East Coast.

In the aftermath of the Sunshine Skyway collapse,  the federal government and the professional  

community, both from the engineering and maritime  sides, invested a serious amount of time and  

investigation into the issue. One culmination was  updated bridge codes that included requirements  

for consideration of vessel collisions. For  highway bridges in the US, those specifications  

are put out by an organization called the American  Association of State Highway and Transportation  

Officials (or AASHTO), but there are similar  requirements worldwide, including in the Eurocode.

A lot of infrastructure is designed for  worst-case scenarios, but at a certain point,  

it just isn’t feasible. This is something  I’ve talked a lot about in previous videos:  

you have to draw a line somewhere that  balances the benefits and the costs. If  

the code required us to design bridges with  Armageddon meteorite or Godzilla protection,  

we just wouldn’t build any bridges. It would be  too expensive. And that’s kind of true for ship  

collisions too. The mass and kinetic energy  of the cargo vessels today is tough to even  

wrap your head around. We just couldn’t afford  to build bridges if they all had to be capable  

of withstanding a worst-case collision. Instead,  for what engineers call “high consequence, low  

probability” events, codes often set the standard  as some acceptable amount of risk. There’s always  

going to be some possibility of an event like  this, but how much risk are we as a society  

willing to bear for the benefit of having easy  access across navigable waterways? In the U.S.,  

that answer, at least according to AASHTO for  critical structures like the Key Bridge, is  

0.01 percent probability in a given year. For some  perspective, it’s roughly the chance of rolling a  

Yahtzee (five-of-a-kind) in a single throw. But  it’s an annual probability, so you have to roll  

the dice once every year. If you did it forever,  it would average out to once every 10,000 years,  

but that doesn’t mean it couldn’t happen twice  in a row. So an engineer’s job is to design  

the structure not to survive in a worst-case  scenario but to have a very low probability of  

collapsing from a vessel impact. And there’s  a lot that goes into figuring that out.

This is the general formula for the annual  probability of bridge collapse due to a ship  

collision. You have all these factors that get  multiplied together. The first one is just the  

number of ships that pass under the bridge in a  year. And there’s a growth factor in there for how  

that number might change over time. Then there’s  what’s called the probability of vessel aberrancy;  

basically, the chance that one of those ships  loses control. AASHTO has some baseline numbers  

for this based on long-term accident statistics in  the US, and the designer can apply some correction  

factors based on site-specific issues like water  currents and navigation aids. Then, there’s the  

geometric probability of a collision if a ship  does lose control. When a vessel is aberrant,  

you don’t know which way it’s going to  head. This gets a little complicated,  

but if you’re familiar with normal distributions  it will make perfect sense. You can plot a normal  

distribution curve centered on the transit path  with one standard deviation equal to the length  

of the aberrant ship to give you an approximation  of where it might end up. The area under that  

curve that intersects with the bridge piers is the  probability that the ship will impact the bridge  

if it loses control. And this is really the first  knob an engineer can turn to reduce the risk,  

because the farther the piers are from  the transit path, the lower the geometric  

probability of a collision. And this factor can  be modified if ships have tethered tugs to assist  

with staying in the channel, something that  wasn’t required in Baltimore at the Key Bridge.

But, even if there is a collision, that doesn’t  necessarily mean the bridge will collapse. This  

is where the structural engineering comes into  play. The probability of collapse depends both  

on the lateral strength of the pier and  the impact force from the collision. But,  

that force isn’t as simple as putting a  weight on a scale. It’s time-dependent,  

and it varies according to the size and type  of vessel, its speed, the amount of ballast,  

the angle of the collision, and a lot more.  Usually, we boil that down to an equivalent  

static load. And based on some testing,  this is the equation most engineers use.  

It’s just based on the deadweight tonnage  (basically how much the ship can carry) and  

its velocity. It’s interesting that they settled  on deadweight, which doesn’t include the weight  

of the ship itself. But again, this analysis is  pretty complicated, especially because you have  

to do it for every discrete grouping of vessel  size and bridge component, so some simplifications  

make sense, and since this one assumes every ship  is fully loaded, it’s relatively conservative.

Just for illustration, the ship that hit the  Sunshine Skyway Bridge had a deadweight of  

34,000 tonnes. The NTSB report doesn’t estimate  the speed at which it hit the bridge, but let’s  

say around 5 knots. That would be equivalent to  a static force of around 56 meganewtons or 13  

million pounds if the ship was fully loaded, which  it wasn’t (but there’s no way to account for that  

in this equation). The Dali has a deadweight of  117,000 tonnes and was traveling at roughly 5  

knots on impact. That’s equivalent to more  than 100 meganewtons or 24 million pounds,  

again, assuming the ship was fully loaded (which,  again, it wasn’t). But you can validate this with  

some back-of-the-envelope physics. Force  is equal to mass times acceleration. We  

know the mass of the ship and its cargo from  records: about 112,000 metric tons. To decelerate  

that mass from 5 knots to a standstill over  the course of, let’s say, 4 seconds requires,  

roughly, a force of 72 meganewtons or 16  million pounds. Even as a rough guess,  

that is a staggering number. It’s 5 SpaceX  Starships pointed at a single bridge pier.

Designing a bridge to handle these forces  is obviously complicated. It’s not just the  

pier itself that has to survive, but every  element of the bridge along the entire load path,  

including the foundation, and (assuming the pier  isn’t perfectly rigid), the superstructure too.  

Again, it’s not impossible to design, but it  gets pricey fast, which is why designers have  

more knobs to turn to meet the code than just  the strength of the bridge itself. One of those  

knobs is pier protection systems. Fenders can be  installed to soften the blow of a ship impact,  

but for ships of this size, they would have to  be enormous. Islands can be built around piers  

to force ships aground before the hit the bridge.  But islands create environmental problems because  

of the fill placed on the river bottom, plus they  get really big for deeper channels, so the bridge  

span has to be wider to keep the channel from  being blocked. Islands can even affect currents  

in the water and the bridge structure, creating  additional load on the foundation as they settle  

after construction. Another commonly used  protection structure is called a dolphin. This  

is usually a circular construction of driven sheet  piles, filled with sand or concrete. Dolphins can  

slow a ship, stop it altogether, or redirect  it away from critical bridge elements like  

piers. The new Sunshine Skyway Bridge used islands  and dolphins to protect the rebuilt span, and  

actually, the Key Bridge had four dolphins, one on  either side of each main support. Unfortunately,  

because it came at an angle, the Dali slipped  past the protection when it lost control.

It’s important to point out that everything  I’ve discussed is a modern look at how engineers  

consider vessel impacts to bridges. When the  Francis Scott Key Bridge was finished in 1977,  

there were no requirements like this, and  the bridge never had a major rehabilitation  

or repair that would have triggered adherence  to the newer codes. Container ships the size  

of Dali didn’t even exist until around  2006. And we don’t know what the ships  

of the future will look like. It’s easy  to say with hindsight that a bridge like  

this should have been better protected against  errant ships, but if you say it for this one,  

you really have to say it for all the bridges  that see similar maritime traffic. And that  

represents an enormous investment of resources  for, potentially, not a lot of benefit to the  

public, given how rare these situations are.  That’s not me saying it shouldn’t be done;  

it’s just me saying that a decision like that  is a lot more complicated than it might seem. I  

don’t expect we’ll see bridge design code changes  come out of this event, but vessel collisions will  

certainly be on the minds of the designers for the  replacement in Baltimore. I’ll let Rob explain.

When you take a look at photos of the Key Bridge,  it looks like Maryland was doing a good job taking  

care of their bridge. So if the NTSB report comes  back and says the bridge was in good shape, it's  

100% the ship that's at fault, well, I don't think  any of us are going to be really that shocked.

But for the old Gerald Desmond Bridge here  in Long Beach, that used to be right here,  

well, the environmental impact report, where they  studied to build this new replacement bridge, the  

port staff really didn't seem too concerned about  a maritime navigation failure. Of a structural  

failure? Let's just say engineers scored bridges  out of 100 points. So you have a brand new bridge,  

it gets 100 points. On that scale, the old  Gerald Desmond Bridge that was right here  

scored a 43. I mean, anything below 80 points,  you get federal money to work on the bridge  

to try to rehab it and get it back into good  shape. And anything, anything under 50 points,  

it's so bad the federal government starts throwing  money in trying to help you replace the bridge.

That's how bad off the Gerald Desmond Bridge was.  Salt from the air of the sea and decades of it  

sitting above sea water and all of that, just nice  salt in the air, eating away at the paint. Well,  

that paint was rated very poor on the  old Gerald Desmond Bridge. And, you know,  

paint protects all the bridge members, all the  metal from rusting out. And as Grady points out,  

every single member of a truss is  really important if you, you know,  

want the bridge to stay in good  shape and not fall down, right?

Engineers also conducted a load analysis.  They tested to see as trucks drove over it,  

how the bridge was holding up. And they found  members of the arch main span were overstressed  

for all design trucks. So that didn't matter  if you drove a big truck or a little truck.  

They were all causing problems with the bridge.  And the concrete that those trucks drove on? It  

was all cracking up. It was rated critical. The  port had to install big nets to catch big chunks  

of pavement that were falling off the bridge  and could hit somebody on the head down here.

So, Long Beach had four objectives that this new  bridge needed to meet in order to build it. And  

those goals may mirror some of the ones Baltimore  may want to have when building their replacement  

bridge. 1. This bridge had to have a design  life of 100 years. Say, stay structurally sound  

for that long. 2. Long Beach wanted to reduce the  approach grades on both sides, even getting up to  

155ft before. Sometimes you were driving up a 6%  grade and now that this thing is over 200 ft high,  

that would be way too steep. So they instead  built these huge freeway viaducts that go on  

and on for like a quarter of a mile to lift people  and trucks gently up to that new bridge height.  

Baltimore's bridge already has some very long  approaches to it, so I don't know whether they're  

going to replace the, uh, ramps approaching the  bridge or not. It'll be interesting to see what  

they end up deciding to do. 3. Provide sufficient  roadway capacity to handle future car and truck  

traffic. The old bridge here was two lanes  in each direction. Four lanes. This widens  

it to six. The Key bridge in Baltimore was also  only four lanes. But this bridge handles twice  

the traffic every day. You know, compared to  the key bridge back when it was open, right?

And both the Key Bridge and the old Gerald  Desmond Bridge had no shoulders for emergency  

vehicles and stalled cars to pull off to the  side. And as you can see, the new bridge has  

these excellent shoulders on both the outside  and the inside of travel lanes. So that makes  

the road a lot safer, because you're not going to  run into the back of a stalled truck in the dark.

It's also a lot safer if you're not in your  car, because this bridge has a way you can  

cross it without being in a car. They've added  this 12ft wide pedestrian and bicycle pathway,  

which is about 12ft wider than what they  had before. Used to be on the old bridge,  

The only way across was inside a car. It's a  good start, certainly not perfect. Right now,  

the path just hits this gate and stops. The city  of LA owns the next harbor bridge down that way.  

It's called the Vincent Thomas Bridge. It's also  old, so it doesn't have a pedestrian walkway,  

so this pathway sort of just ends at the  city limit because there's nowhere for it  

to go. But adding in a multi-use path like  this one onto the new Key Bridge would be  

such a no-brainer. It could take a four mile bike  ride from like, Hawkins Point to Sparrows Point,  

down from four miles with the bridge  path from 22 miles right now, without it.

And finally the fourth goal: providing vertical  clearance for new generation of larger vessels,  

which the new bridge certainly has.  And that must make the port very happy.  

And I'm willing to bet that Baltimore will  take that goal and maybe turn it on its side  

and talk about horizontal clearance and insist on  a design that eliminates the risk of an allision,  

like what happened on March 26th  from ever happening again, Grady.

Thanks Rob. If you love deep dives  into transportation infrastructure,  

go check out his channel after this. But,  it’s important to point out that this wasn’t  

just a bridge failure; it was a bridge failure  precipitated by a maritime navigation failure.  

Obviously, engineers who design bridges  don’t have a lot of say in the redundancies,  

safety standards, and navigation requirements  of the vessels that pass underneath them. But  

if you look at the whole context of this tragedy  and ask, “How can our resources best be used to  

prevent something like this from happening  again?”, reducing the probability of a ship  

this size losing control has to be included with  the structural solutions like pier protection  

systems. I don’t know a lot about that stuff,  so I couldn’t tell you what that might include,  

but I’m sure NTSB will have some recommendations  when their report eventually comes out. Having  

tugs accompany large ships while they  traverse lightly protected bridges seems  

like a prudent risk reduction measure, but  that’s just coming from a civil engineer.

And, speaking of risk reduction, I have to say  that using risk analysis as a tool for design  

is really not that satisfying. We humans are  notoriously bad at understanding probabilities  

and risks, and engineers are not that great  at communicating what they mean to people  

who don’t speak that language. That’s how we get  confusing terminology like the hundred-year flood.  

And it’s unsettling to come face-to-face with  the idea that, even if our bridges are designed  

and built to code, there’s still a chance of  something like this. Everything’s a tradeoff,  

but the people driving over the bridge (or working  on it) had no direct say in where the line was  

drawn or whether it applied retroactively, even as  ships got bigger and bigger. But I hope it’s clear  

why we do it this way. The question isn’t “Can we  design bridges to be safer?” The answer to that  

is always “yes.” The real question is, “How much  risk can we tolerate?” or put in a different way,  

“How much are we willing to spend on any  incremental increase in safety?” Because the  

answers to those questions are much more complex  and nuanced. And if all bridges were required  

to survive worst-case collisions with ships like  Dali, we would just have a lot fewer bridges. But  

sometimes it takes an event like this to remind  us that risks aren’t just small numbers on a piece  

of paper. They represent real consequences, and  my heart goes out to the families of the victims  

affected by this event. I hope we can honor them  by learning from it and making improvements,  

both to our infrastructure and our maritime  systems, so that it doesn’t happen again.

The Key Bridge collapse is a really interesting  case study because it combines two really  

different areas of understanding. Obviously,  I can approach it from the bridge engineering  

perspective, but there’s just as much, if  not more, to understand from the maritime  

side. Ships and marine navigation are  literally a whole different world,  

and my friend Sam from Wendover Productions  has an awesome video specifically about  

the crazy sophistication of the commercial  fishing industry. I find it so fascinating  

to learn about the little details behind  industries I know nothing about, so honestly,  

this entire series, “The Logistics of X” is  so good, and it is only available on Nebula.

You probably know about Nebula now, even if  you’re not subscribed. It’s a streaming service  

built by and for independent creators. No studio  executives deciding what gets the green light,  

no advertisements, and no algorithm driving  the content into a single style. It’s just  

independent creators making stuff their excited  about with a few barriers and distractions as  

possible between you and us. The website just  got a huge update that completely redesigned  

the home page, making it easier to find  new stuff in addition to your favorites.

My videos go live on Nebula before they come  out here, and my Practical Construction series  

was specifically produced for Nebula viewers who  want to see deeper dives into specific topics.  

I know there are a lot of streaming platforms  out there right now, and no one wants another  

monthly cost to keep track of, but I also know  that if you’re watching a show like this to end,  

there is a ton of other stuff on Nebula that  you’re going to enjoy as well. So I’ve made  

it dead simple: click the link below and you’ll  get 40% off an annual plan. Pay just one time,  

30 dollars, for an entire year’s access at  nebula.tv/practical-engineering. Or if you have  

subscription fatigue, but still want to support  what I’m doing, you can get a lifetime membership.  

Pay once and have access for as long as you  and Nebula last. Hopefully that’s a long  

time! If you’re with me that independent  creators are the future of great video,  

I hope you’ll consider subscribing. Thank you  for watching, and let me know what you think!