Francis Scott Key Bridge Collapse - Baltimore, MD

[Arthur Dent]

Echoing what you said, it seems very odd that there is a single point of failure (in this case electrical breakers HR1 and LR1 opening/tripping) which caused loss of power to the cooling pumps which caused the main engine to shut down. You'd think everything critical to the engine (eg, the pumps) would have an emergency backup, like less than 10 seconds. And, that would either prevent the engine from shutting down completely or would allow it to be functional within a very short time, again, thinking less than 10 seconds.

Yeah, that seems very odd. It sounds like the breakers on the transformer tripping is an uncommon failure mode, but they had a second redundant transformer available. Like you say, why would there not be a mechanism to automatically connect it in time to keep the engine going or at least quickly restarted?

Seems like there's a lot we don't know. Why did the transformer breakers trip? Why didn't they automatically switch to the backup transformer? Report says they restored power the first time by manually resetting the transformer breakers that tripped, but, naively, with a full backup transformer available, it seems like it would make more sense to try that one before retrying the one that just failed. Then why did the big generators fail together but in a way that another parallel standby generator could still start? They say the main engine relied on various fuel pumps, etc running off the LV bus. Is the same true for the generators? Maybe the loss of LV bus power caused some issues with them that only manifest after a delay? The report says the crew switched in the second transformer after the second blackout but never mentions if the breakers on the first transformer opened again. A footnote says that after the accident, they set the switches to continue powering the ship via the standby generator 2 that came on automatically after the second blackout and only the second transformer, so it could be that the breakers on the first transformer were opened manually after the accident, but if they had to manually connect the second transformer to restore power after the second blackout, the first transformer must not have been working if it was still connected.

The story about the day before blackout is also interesting in that it seems like it is another instance where fuel pressure to the generators fails immediately after a blackout. The mechanism of switching in a backup generator does not seem to go smooth in general.

I guess my reading is that the initial cause of of failure is some kind of fault in the transformer and then the rest played out because the mechanisms to respond to such a failure were either no good or not properly implemented, but it's all pretty vague.

Kind of the worst case scenario unfolded. The bridge wasn't protected, the ship lost power at exactly the wrong time, and the conditions dictated the course from there which unfortunately was right into the unprotected pier.

According to this video, it's basically hopeless if you have a blackout in a situation where you need to maneuver. Even in training simulations, a blackout almost almost always leads to a crash.

https://youtu.be/jxeKXjDVqMA?si=b_qWw8Rk5yOEgX6w

Which seems kinda of crazy. If a blackout is so devastating, you'd think you want to try and create some kind of emergency system to give minimal emergency maneuvering. Or, at least, more should be done to reduce single points of failure and/or give greater odds of rapid power restoration. Is that totally cost prohibitive, or has it just not been taken seriously?

Anyway, I enjoy thinking about these things and having someone to discuss it with and did not mean to come off like I was blasting your ideas. If I did that, I'm sorry.

Hopefully, I will not forget about it all while we wait the year it seems like it's going to take to get the full story.

[AWvsCBsteeeerike3]

@Arthur Dent

All fair questions regarding why the transformers tripped and how it was handled after that. I'm out of my element when we get into the circuits, transformers, breakers, and switches.

My only knowledge about these things, aside from college which is all pretty useless at this point, is working on a couple hospital projects with a large MEP firm that was replacing all the individual generators at individual buildings with a generator plant. This allowed the hospitals to save substantially on their electric bill (reduced rates if the hospital agrees to run the generators for X hours a year). We ended up rerouting the primary to the generator plant and using the plant as the main switch. From there, normal and redundant essential lines were separately run (different conduits, same duct bank encased in concrete) throughout the campus. And, the plant could provide the voltage required via the primary power feed or by running the generators.

The essential loads were different than the normal loads iirc with essential only dealing with critical life safety equipment (from memory but don't quote me it was stuff like equipment in Operating Rooms, backup lighting, ventilators, monitors, etc). And in the event of a primary power outage (grid power outage), the generators could energize the essential power lines within I want to say 10 seconds, hence why I quoted that earlier, with the normal lines being energized within 45 seconds to a minute.

Following that line of thinking, it would make sense to have an 'essential' bus or busses to keep the critical items (pumps, whatever the main engine needs, steering, navigation) going at all costs. Especially in critical situations, like in port. Sounds like they have 4 generators plus an emergency generator? Why isn't one of them (let's say Gen 3 in this case or Gen 3 and Gen 4 if required) running on standby, and in the event Gen 1 and 2 go offline causing a blackout, there's an immediate switch to the essential busses from Gen 3/Gen 4. Seems like this could be automated and done within seconds but I don't know for certain. I can't imagine the cost for this is much, relatively speaking. From my understanding, you're just talking about some gear in the mechanical room and using the equipment already on board along with the fuel to run the generators on standby for an hour or two.

Anyway, I enjoy thinking about these things and having someone to discuss it with and did not mean to come off like I was blasting your ideas. If I did that, I'm sorry.

Me too. No harm, I had zero confidence in what I was saying about the maneuvering but enjoy thinking about what could have happened. i am now even more curious as to what could have been done had they got the main engine restarted within 30-45 seconds and what actions they would have taken.

And, yes, unfortunately by the time the final report comes out, I will have forgotten everything.

[AWvsCBsteeeerike3]

Looking a little more into it, emergency power systems for hospitals falls under the National Fire Protection Agency code 99 and 110 which requires critical and life safety branches to have power restored within 10 seconds. Apparently this is where the diferent types of generators come into play as some are classified to do this. And, there's more cost associated with faster generators.

So it may be that ships generators, switches, load terminals, etc aren't capable of restoring power this quickly. That said, I'd have to think having a generator running in standby would decrease the time, and upgrading the switches/terminals/whatever if necessary would get you to a more manageable timeframe of power restoration.

[Arthur Dent]

Following that line of thinking, it would make sense to have an 'essential' bus or busses to keep the critical items (pumps, whatever the main engine needs, steering, navigation) going at all costs. Especially in critical situations, like in port. Sounds like they have 4 generators plus an emergency generator? Why isn't one of them (let's say Gen 3 in this case or Gen 3 and Gen 4 if required) running on standby, and in the event Gen 1 and 2 go offline causing a blackout, there's an immediate switch to the essential busses from Gen 3/Gen 4. Seems like this could be automated and done within seconds but I don't know for certain. I can't imagine the cost for this is much, relatively speaking. From my understanding, you're just talking about some gear in the mechanical room and using the equipment already on board along with the fuel to run the generators on standby for an hour or two.

It's weird. They did have an emergency backup generator that powers a limited emergency bus, and it did come on after the first blackout, though it seems like it took a full minute, which seems to more than the mandated 45 seconds. But it doesn't seem like that time loss is the main issue because the main engine shut down almost immediately, and would need to be restarted regardless. It's not clear to me exactly what the emergency bus powers either. We know it powers an emergency pump to control the rudder, but it seems like that's almost pointless as the rudder did almost nothing without the engine running. I would like to know if it powered systems needed to keep the big generators running, as there is no explanation why these failed some two minutes after the initial blackout.

And per your thinking, it sounds like they did have this standby big generator auto start when the primaries went down, but it just didn't make a difference.

It seems like they already basically have all the redundant expensive gear needed to quickly react to a failure, but the control systems or operational practice make the redundancy ineffective or, at least, too slow.

According to this video, there is actually a fairly simple approach that would likely have prevented this: opening the bus split breakers and running both sides in parallel when in a maneuver critical environment. This way, a failure would only blackout one side or another, which would leave enough equipment online to avoid an immediate disaster. Sounds like it's a no one fixes the problem until after the disaster situation more than a it's just way too expensive problem.

https://youtu.be/zGi7BBhanIs?si=SZyB9WmhbINY1cPO

[G. Keenan]

So many missed busses . . .

[AWvsCBsteeeerike3]

As he notes, ships, just like anything else being built/operated, are going to follow the regulatory codes, regulations, et al.

Little out of my element here but... Looking at the diagram in the video connecting the busses so they're tied together makes a lot of sense and would provide the redundancy that could have prevented this. If I understand correctly, had Gen 1 and Gen 4 been operating and the busses (6600V, 440V at least) been tied together, the opening of HR1 and LR1 would not have caused a blackout since Gen 4 and TR2 would have still provided power to everything. The HV bus is directly tied together with no transformer needed and the 440V would have been supplied through transformer 2. Not sure if that is true or not, but looks like it would based on the diagram. If so, for shame. I don't know what additional costs there are doing it that way, but seems like a pretty straightforward approach without too much of a financial burden.

Something that shouldn't be lost in all of this conversation is just how ineffective the rudder was once propulsion was lost. Think I mentioned this earlier, but it makes sense. Rudders act as wings in water. But the prop is directly in front of the rudder so the speed of water over the rudder is much greater than the ships speed through the water. Without the increased flow from the prop, the rudder is akin to a stalled airplane wing.

I guess it goes without saying that if a ship loses propulsion it's in big trouble. Still, having driven recreational boats powered by outboards for the past 20 years or so, I found it interesting how the rudder had so little impact.

In many ways, this was an unfortunate series of events. At the same time, it's not unfathomable to imagine a similar, much more tragic event unfolding given the same setup.

[Arthur Dent]

Little out of my element here but... Looking at the diagram in the video connecting the busses so they're tied together makes a lot of sense and would provide the redundancy that could have prevented this. If I understand correctly, had Gen 1 and Gen 4 been operating and the busses (6600V, 440V at least) been tied together, the opening of HR1 and LR1 would not have caused a blackout since Gen 4 and TR2 would have still provided power to everything. The HV bus is directly tied together with no transformer needed and the 440V would have been supplied through transformer 2. Not sure if that is true or not, but looks like it would based on the diagram. If so, for shame. I don't know what additional costs there are doing it that way, but seems like a pretty straightforward approach without too much of a financial burden.

I could imagine some reasons why it might be a problem to do that compared to split bus parallel operations. For example, the two transformers may not be perfectly matched which may cause overloading circular currents, and maybe various types of faults could end up cascading if everything is locked together, but I'm just speculating.

Anyway, since they do have this redundancy, and you'd only really need parallel operation in the tiny fraction of the voyage near port, it really seems like there should be methods to use redundancy to dramatically reduce the chance of a black out induced catastrophe at quite modest cost. It will be very interested to see what the NTSB ends up recommending.

Something that shouldn't be lost in all of this conversation is just how ineffective the rudder was once propulsion was lost. Think I mentioned this earlier, but it makes sense. Rudders act as wings in water. But the prop is directly in front of the rudder so the speed of water over the rudder is much greater than the ships speed through the water. Without the increased flow from the prop, the rudder is akin to a stalled airplane wing.

I guess it goes without saying that if a ship loses propulsion it's in big trouble. Still, having driven recreational boats powered by outboards for the past 20 years or so, I found it interesting how the rudder had so little impact.burden.

On the one hand it makes logical sense, but at the same time it feels shocking how utterly useless it is. I assume it's a much more severe loss of steering without the motor on a big ship than on a recreational boat? I guess it must also be related to rudder size as, obviously, you have effective rudders on sailboats, barges, and whatnot with no motor, which I'm guessing is just done by making it bigger. There's clearly some kind of regulation requiring emergency rudder control when the motor is out, and I wonder whether the authors propagated that rule from smaller boats where this works better or what.

[AWvsCBsteeeerike3]

@Arthur Dent

Makes sense about issues syncing the currents from the transformers. Thanks. But, yeah, seems like they have everything they need and surely there is a way to effectively eliminate the loss of critical equipment in critical situations.

haven't been able to find out a lot about the rudder Dali had or how exactly they're designed. There are multiple types and some (flap rudders) have a second pivot point in the rudder itself on top of the pivot about the hull. These allow for more curvature of the rudder enhancing the physics behind them (again, going to aeronautics, like when flaps are extended for landing).

Outboards and inboard/outboards on recreational boats are a different setup than a prop/rudder setup like exists on recreational inboards as well as in this case larger container ships. In the former, the prop is aftmost part of the lower unit typically with a small 'rudder' beneath it. The entire lower unit rotates so not only is the 'rudder' turning but also the prop is turning which changes the angle of propulsion/force so you don't really need the rudder for directional control since the direction the force is being applied can be changed. Not sure if they call the little tail a rudder or not on lower units. But, even when the prop isn't spinning (such as the case with pulling up to or into a dock), the angle of the lower unit will still provide some directional control.

Sailboats have rudders, still usually behind the prop, but they work even when the props not spinning by deflecting water moreso than creating a velocity induced pressure difference about the rudder that will create lift or rotational forces. I'm sure it still happens, but that pressure difference is negligible, from my understanding, compared to the force of the water hitting the rudder.

And, this where I get fuzzy.

In larger vessels like the Dali and with more complicated rudder designs, the prop is directly in front of the rudder. When designing the rudders, how much are they relying on a prop induced increased velocity of water hitting the rudder. Obviously a lot. They're not designing for sailing. And when doing so, how big of a factor does bernouilli's principle play in it compared to the more obvious/simple forces of water hitting a rudder? I'd guess it's negligible in sailboat rudders. Given the evolution of ships' rudders, it's apparent that it's not negligible in those calculations. Either way the deflectional forces are a large part which will be diminished by the loss of velocity associated with a dead engine, but they'll still be somewhat present unless the speed and direction of the vessel over ground ground is the exact same as the current which is highly unlikely. But the bernouilli's principle forces will be completely erased.

What we saw with the Dali, the rudder was almost (completely?) useless.

Not sure if that makes sense or answered your question.

Here's a quick video going over the different types of rudders. Not very empirical but at least schematically correct.

https://youtu.be/jWD-IxjdFrs?si=etlu6ccgzUU1UhF8&t=41

[AWvsCBsteeeerike3]

Regardless, they're not going to change rudder design to account for a complete failure of the engine. No more than they would require Boeing 737 Maxs to have big ole parachutes to account for the faulty MCAS.