All Articles Infrastructure Construction Q&A: Considerations for the Key Bridge replacement

Q&A: Considerations for the Key Bridge replacement

The collapse of the Francis Scott Key Bridge prompted much debate in civil engineering circles. Here's what an AASHTO committee member has to say.

9 min read


David Adams/Wikimedia Commons

On March 26, an unspeakable tragedy occurred as a large vessel struck Baltimore’s Francis Scott Key Bridge and caused its collapse and the presumptive death of six people. The catastrophic failure has spurred wide debate in civil engineering circles regarding whether more could’ve been done to prevent the bridge’s collapse and what kind of structure should replace it. In a poll of those who subscribe to SmartBrief for Civil Engineers, 37% of respondents indicated better operational guidance, such as rules for tug boats and cargo vessels, would be the best preventive measure for a similar future event. Thirty-six percent suggested better bridge protections, such as fenders and warning systems, should be the primary preventive measure. Twenty-one percent said both operational and bridge protections should be implemented. Only 5% said nothing could’ve been done, and no one suggested designing the bridge to a greater design strength. One point that has been brought up repeatedly, notably by NTSB chair Jennifer Homendy, is that there was no redundancy in the bridge design. Dr. Abi Aghayere, a professor at Drexel University, told SmartBrief that lack of redundancy explains “why the bridge went down so quickly in a non-ductile or brittle manner.” 

Online, some civil engineers have suggested the tragedy will lead to a change in standards by the American Association of State Highway and Transportation Officials (AASHTO) – the organization that, among many things, publishes and updates the standards that allow engineers to design bridges for protection from ship collisions. To learn more about the design considerations for a future replacement to the Key Bridge, we spoke with Jason Hastings, chief of bridges and structures at the Delaware Department of Transportation and vice chair of the AASHTO Committee on Bridges and Structures. The following interview has been edited for clarity and brevity. 

SmartBrief: A civil engineering professor at Johns Hopkins University has suggested that the tragedy should be seen as “a full infrastructure problem, as opposed to a bridge problem.” Would you agree? Where do you stand in the chicken-and-the-egg debate that has arisen over where to focus attention? 

Hastings: I think I agree. Obviously it was the bridge that had the ultimate collapse, but the vessel had a failure of its own – a loss of power. The initial investigation shows that ultimately led to the ship colliding with the pier. NTSB will be doing a thorough investigation and will sort through all of the details and ultimately come up with recommendations for various agencies or entities to hopefully improve upon what led to the ultimate cause of failure.

SB: Many civil engineers have speculated that the collapse will prompt a change in AASHTO standards. What can you share about conversations among the AASHTO Committee on Bridges and Structures since the tragedy, and what standard changes do you envision? 

Hastings: It’s hard to envision anything specific at this point. Certainly, we will take into consideration the changes in various conditions that have occurred over the years since AASHTO’s original vessel collision specifications were developed. The first guide spec was published in 1991, and then from a formal design specification, that was included in the LRFD bridge design specs in 1994, which was the original edition of those. We’ll take a look at what changes have occurred in vessel sizes and about the loadings since then, but it’s important to note that every bridge is designed a little bit differently because it’s site-specific. We’re not going to design a bridge over a highway for a vessel collision. So bridge owners have to keep that in mind. Ultimately, the design is based on specific site conditions. The vessel collision provisions in the design spec sort of lend that way. They give a framework for a minimum design, but ultimately, the owner of that bridge needs to understand what kind of vessels are navigating through that channel and take into account the specific loading that could occur based on the vessel traffic. Ultimately, our design specs are based on probability. Hopefully, it’s not a surprise to anyone that there is a probability of failure, even if it’s obviously very, very low. We can’t design an infinite-life bridge because it would be too costly. The bridge would be too massive and just create too much of an impact on the environment. All of our design codes are calibrated for some level of reliability, but that reliability does have a very, very small probability of some level of failure.

SB: Have you met as a committee since the March 26 tragedy? 

Hastings: We have not. I mean, just a couple of folks on the AASHTO Executive Committee have chatted by email with us a little bit, just in coordinating discussions with the media, just to make sure that we’re answering any questions people have. We do have our annual meeting coming up. And obviously, this will be one of many topics of discussion. I’m certainly looking forward to whatever the NTSB or FHWA has to offer in terms of recommendations as NTSB’s investigation moves ahead.

SB: In 1977, when the Key Bridge opened, fenders existed, but were not yet standard. Three years later, the Sunshine Skyway Bridge in Tampa Bay collapsed after a freighter hit a support pier during a storm. It wasn’t until the 90s that AASHTO required devices like fenders wherever bridges are not “designed to resist vessel collision forces.” Should the Key Bridge have been retrofitted to meet the 1994 standard? 

Hastings: Ultimately, I think that that question lies with the bridge owner. It’s the bridge owner’s responsibility to see if there are any changes in condition. So, if all of a sudden, a new marina moves in next to a bridge, that’s going to change the type of vessel traveling on the waterway. It would be on the owner to ask engineers to see if it changes anything with the structure. We would do the same thing if there was some change in truckloads. There can be many different things that would cause the DOT or the bridge owner to have to relook at a bridge. The vessel collision guide spec is pretty clear in how it’s supposed to be carried out for reconstruction or replacement or new bridge rehabilitations. It’s understandable why maybe the bridge owner didn’t look at it. My understanding is the port has been there since before the Key Bridge. So, maybe they’re saying “Okay, well, the port’s active, we have a bridge here that works,” and then some different provisions come along. But as a bridge owner, you’re responsible for thousands or tens of thousands of bridges, all getting routinely inspected and maintenance carried out and a whole host of different things. If something specific isn’t changing with the bridge, it’s likely that it’s not going to be investigated to that level.

SB: According to the Federal Highway Administration, 17,468 of the 615,000 bridges in the United States are fracture critical, meaning that if one piece of the bridge support fails, the whole bridge could collapse. After the collapse of the Francis Scott Key Bridge, how do you expect those fracture-critical structures will be reviewed? 

Hastings: For fracture-critical breaches, there’s already an additional requirement in terms of safety inspections. Our typical bridge safety inspector has to pass a two-week course and then be refreshed every five years. I think across the country, bridge owners are taking a second look at the navigable waterways and bridges over those navigable waterways, seeing if there are any risks out there. Certainly in Delaware, we’ve taken a second look. In our bridge inspection report, there are different codes, or different elements, and one of them is pier protection. We had a few of our bridges that were incorrectly coded for that, so we’re going to go through and clean up some of those things just to make sure they’re reporting properly. Again, not a risk, but sometimes there are things in there that are like, “oh, we caught this.” We all do our due diligence to make sure that there’s not something that potentially could be a risk within our own state inventories.

SB: A USA Today article recently said “[a]ssuming the Key Bridge is rebuilt, it will likely be a cable-stayed or suspension bridge.” Do you agree, and if so, what is your opinion, independent of AASHTO and DelDOT, on what kind of replacement structure makes the most sense in terms of compliance with modern design? 

Hastings: Well, it’s unlikely that they would put a fracture-critical bridge back. At this point, we typically do everything we can to stay away from a fracture-critical structure. I’m sure the Maryland DOT is going to do their due diligence in terms of structure type, but with an emphasis on speed, getting that bridge back up and running. The span lengths that are needed and the bridge height would certainly lend itself to a cable stay or suspensions. Some sort of bended-type bridge. Definitely a longer span for a high-level bridge.

SB: Would post-tensioning be an option in that environment? 

Hastings: Over a waterway, at least in Delaware, we would tend to lean more toward concrete. That’s not a knock on steel, as there are plenty of places where steel is a good option. But this wouldn’t be a decision right on the coast by the ocean, where there’s a lot of salt spray like what we have in Delaware. So to put a steel bridge back would certainly not be surprising. But for a concrete option, post-tensioning can certainly be a big piece of that, or maybe even a requirement, for upgrades that long.

SB: Last year, the Delaware River and Bay Authority got started on the Delaware Memorial Bridge Protection System, which includes eight solid-fill dolphin cells, each measuring 80 feet in diameter. Four cells installed at the piers would support both eastern and western towers and be located a minimum of 443 feet from the edge of the Delaware River’s channel. Should MDOT consider a similar system for the Key Bridge replacement? 

Hastings: It’s certainly a good option. The vessel collision provisions give the owner quite a few options. They could do dolphins or some sort of protective, sacrificial, element like that. You can always do a rock island so that the ship can’t actually get up to the pier. You can design the pier for the loads, although the loads of the ships that are going through that channel are massive. I mean, that you see the nature of that force that obviously brought down the Key Bridge. It’s just a massive, massive force. So you’re more likely to end up with some other alternative to designing the pier to handle that load. 


If you enjoyed this article, click here to subscribe to SmartBrief for Civil Engineers.