Baltimore's Francis Scott Key Bridge Collapses after Ship Strike

[Vanadium 50]

Let's turn it around. Suppose I can make bridges twice as unlikely to fail by increasing the cost by a factor of 10. Should I?

Might it not make more sense to build more bridges to increase redundancy rather than fewer but stronger bridges?

I don't know the exact number, but there are ~1500 bridges at least that long in the US. When was the last collapse? 2007 and I-35W? So we are talking a MTBF of something like 20,000 years.

To me, this sounds pretty good. Spending a lot of energy trying to get this to 30,000 or 50,000 seems tough. If you want to argue "nothing is more important than safety", by making drivers take the long way, you are causing more automobile fatalities. These are less dramatic and seldom make the news, but that doesn't mean they aren't real. The optimum point may be more, cheaper bridges.

 

[Vanadium 50]

To extend:

Say a bridge carried 10M cars per year, and saves 5 miles per trip. So that's 1T miles saved per bridge. Taking a car fatality rate of 1 per 100M miles, that means a bridge will save 10,000 lives. If you say typically there are ~10 fatalities per bridge failure, that's still 1000x better.

To me that says that the winning proposition is more bridges, not fewer but safer bridges.

 

[nsaspook]

How could they, realistically?

Lots of forces on a bridge. Lots of kinetic energy in a ship. A bunch of hooring and hollering by politicians and talking heads won't change that.

+1
Look at the relative size of ships and the bridge.

The ship coming through the new channel is not exactly a small boat, yet it is completely dwarfed by the Dali.

 

[jack action]

Say a bridge carried 10M cars per year, and saves 5 miles per trip. So that's 1T miles saved per bridge. Taking a car fatality rate of 1 per 100M miles, that means a bridge will save 10,000 lives. If you say typically there are ~10 fatalities per bridge failure, that's still 1000x better.

That only works if people do end up doing 5 miles less overall. This is what hurts this type of theory: people just do the same amount of mileage, just going somewhere else. Worst of all, easier access to the infrastructure usually just lengthens the commute for most people - because they choose to drive the same amount of time - and they end up driving longer distances.

There is no way a bridge will save 10,000 lives.

 

[phinds]

There is no way a bridge will save 10,000 lives.

 

[Vanadium 50]

There is no way a bridge will save 10,000 lives.

It will over a 20,000 year MTBF. That's 2 lives/year/bridge. To me, that sounds about right. If you like, take an order of magnitude off if you like - doesn't change the conclusions.

You are correct that people have already decided what is an acceptable risk. In the US, it;s about 40,000 fatalities per year. (And my estimate is that without bridges it would be 43,000) The number killed in suspension bridge failures is about 1. Perhaps 2. "Heavens to Betsy! Something must be done!"

 

[Frabjous]

If I understand this correctly
https://conference-service.com/pianc-panama/documents/agenda/data/full_papers/full_paper_46.pdf
the US standard for annual frequency of collapse for critical bridges is 0.0001

 

[Vanadium 50]

the US standard for annual frequency of collapse for critical bridges is 0.0001

Or 1/10,000. My estimate was 1/20,000. (We may define "major bridge differently.)

They claim 35 losses worldwide in 55 years. If the US is 10% of the total, that's a MTBF of 24000 years. If instead its 20%, it's 12000 years.

 

[Rive]

That's 2 lives/year/bridge.

I don't really understand that why 'bridge safety' went down only this way, but I think the money lost on this single accident would be a far more fitting explanation for what's expected to follow than the six (?) casualty.

 

[Vanadium 50]

No objection to that, but if you take the cost of replacing a bridge and spread it around to all the bridges, its $100,000/year. That doesn;t buy you much safety. Heck, it doesn't even buy you much in the way of drawings.

I suspect - but do not know - that a purely economic optimization will lead you to more, but flimsier, bridges.

 

[Rive]

I think as safety, for cases like this we are talking about a pile of gravel dumped around the pillars. Or something like that.
And only for cases where big ships are expected. I can imagine they'll just leave the most alone.
Not like a new layer of containment for an NPP or anything. Just a big pile of gravel dumped at the right place.

 

[Astronuc]

How could they, realistically?

Lots of forces on a bridge. Lots of kinetic energy in a ship. A bunch of hooring and hollering by politicians and talking heads won't change that.

Impact/collision analysis, which is feasible. One can do 'what if' scenarios, and follow with probabilistic risk assessment, as well as a cost benefit analysis. One might decide to spend $50 to 100 million to protect a bridge that would require $5 billion to replace (and protect).

An informed risk assessment requires one to acknowledge that risks to change, e.g., with increase in ship size (and mass), such that old analyses from say 50 years ago may not adequately reflect the current risk. Also, part of the risk also involves the frequency in which ships have power and/or navigation/steering failures. It may be less expensive to require ships to ensure their power and navigation systems are fully functional, with appropriate, backup before leaving port. Or perhaps require tugboats must accompany large ships, or ships deemed to have problems prior to passing critical infrastructure.

 

[Rive]

It may be less expensive to require ships to ensure...

Given the number of ships actually it may be more expensive, but with somebody else paying the bill, and that matters

 

[gleem]

Given the number of ships actually it may be more expensive, but with somebody else paying the bill, and that matters

Ultimately the consumer pays the bill either through taxes or by passing through the increased corporate operating expenses.

 

[Flyboy]

I think that people ate trying to generalize a solution to very situationally specific issues. Every bridge is different, from construction to traffic to risks, and no one solution will satisfy every problem.

In this particular case, the electrical problems should have been sorted out before departure. But that’s with the benefit of hindsight and an armchair to speculate from

 

[Nik_2213]

The converse of this event would seem to be protection of Arctic drill-ships from ice-bergs.
IIRC, following fraught learning curve, a 'layered' approach proved necessary.

For 'harbour' bridges, and such as Thames Barrage, stopping any sizeable incoming 'cold' just needs too much infrastructure / mass. Such would require a non-trivial island, retro-fit potentially narrowing shipping channel and raising the risk of lesser, but channel-blocking collisions...

IIRC, by analogy with freeway edge / motorway median barriers that progressively deform to dissipate energy, re-direct / significantly mitigate impactors, bridge piers etc need outer frangible bollards, middle stronger bollards and inner 'hard place'. Traversing these obstacles dissipates enough energy to swing arrivals from wrecking radial to survivable lateral...

Okay, yes, Titanic would have probably survived t-boning that berg, where-as the lonnng flank scrape popped enough flank plates to take them down. But we want the bridge pier to endure...

 

[Baluncore]

The vast majority of the Key bridge was protected by shallow water, where a vessel of sufficient size could not approach the support columns. There were only the diagonal approaches from the main channel left open to a sufficiently large vessel. (It could be said that the Suez Canal banks were well protected from the Ever Given, by the shallower water at the edge of the canal).

The bridge was "probed" on a daily basis over many years, as part of an unintentional Monte Carlo experiment to find weak points in the defences. It was not until a vessel of sufficient size became available, that the vulnerability was identified by destructive testing.

I would be interested to know how the main channel depth was increased over time by dredging, to allow progressively larger vessels to pass. It seems the mindset of increasing channel depth and capacity, opposed the progressive buildup of parallel protection structure, or banks on either side, that would have narrowed the main channel.

The other issue here is the progressive domino failure of the entire support truss. The majority of the investment could have been protected by having a structure that could suffer only partial collapse. It seems to me that the large size of the truss members, precluded redundancy, in the support structure of the main spans.

 

[Swamp Thing]

 

[Borg]

I think that they've been waiting until they recovered the last body - which was recovered this week.

 

[Hornbein]

 

[Borg]

Explosive charges are scheduled to be set off at 5pm EST today in just under an hour.

Live view here:

 

[Borg]

The video seems to not be live since it did go off at 5 but the video still shows it not exploded 10 minutes later.

 

[nsaspook]

 

[Astronuc]

Cargo ship that caused Baltimore bridge collapse had power blackout hours before leaving port
https://apnews.com/article/baltimor...-report-ntsb-b6d4797441350823adb7721164848c04

According to the preliminary report, at 1:25 a.m. on March 26, when the Dali was a little over half a mile away from the bridge, a primary electrical breaker that fed most of the ship’s equipment and lighting unexpectedly tripped, causing the ship to lose electrical power and experience a blackout. The main propulsion diesel engine shut down after the pumps lost electrical power. The ship’s crew was able to restore power, then called for an assist from tug boats and the senior pilot ordered the ship’s anchor to be dropped.

A second blackout then occurred and a marine radio call was made to warn waterborne traffic. The ship then struck a main support pier on the bridge, causing it to collapse within seconds.

 

[DaveE]

NTSB preliminary report is out. Lots of information in this video, but no ultimate answer to why. Apparently an issue with electrical distribution, not engines or generators.

 

[Astronuc]

Eventually, one will find a report here - https://www.ntsb.gov/investigations/Pages/DCA24MM031.aspx

https://www.ntsb.gov/investigations/Documents/DCA24MM031_PreliminaryReport 3.pdf

 

[Tom.G]

The report requires a PASSWORD to open!

UPDATE: It depends on which PDF reader is used, see 2 posts down.

 

[DaveE]

The report requires a PASSWORD to open!

Not for me. Try again?

 

[Tom.G]

Adobe Reader opens it fine, Nitro Reader wants a password.

Thanks @DaveE.

[edit] Interesting/informative read.

 

[berkeman]

https://www.ntsb.gov/investigations/Documents/DCA24MM031_PreliminaryReport 3.pdf

Interesting how the "Dolphin" protection system did not help because of the track that the Dali took. I wonder how much that Dolphin that is in-line with the pier would have helped if the Dali were on a straight course for the pier.

Also, an interesting picture of the NTSB investigator applying anti-tamper materials to the electrical panels...

 

[Astronuc]

Report on Key Bridge collapse raises questions with potential legal ramifications
https://www.yahoo.com/news/report-key-bridge-collapse-raises-194800100.html

BALTIMORE — The crew and pilots aboard the massive container ship that destroyed the Francis Scott Key Bridge weren’t under the influence of alcohol or drugs.

Nor was the ship running on contaminated fuel, which some experts postulated as a cause for the 984-foot Dali losing power twice as it approached the span in the early morning of March 26.

But the revelation from federal investigators that the Dali experienced two complete power outages about 10 hours before its ill-fated departure raised new questions about how the vessel’s crew responded to the electrical failures, who was informed of the blackouts — and whether the vessel should have set sail.

 

[Vanadium 50]

So, how long after fixing a power failure do we wait before moving? If 10 hours isn't enough, what is the right number? (Keeping in minf that conditions in port are not the same as conditions at sea.)

 

[Rive]

By 'fixing a power failure' you mean getting the power back or ensuring that everything is really up to code?

 

[Vanadium 50]

ensuring that everything is really up to code?

I don't know what that means.

 

[Rive]

?

For every complex (electric and other) system has regulations how it is built, how it is operated and such. That's what makes 'safety' more than just a word.
If there were outages on that ship then somewhere, something was not up to code.
Restore power is just one thing. Finding the root of the trouble and bring it back up to code is what 'fixing' should mean, at least by my humble opinion.

I think the remains of that bridge would agree