what is the inside scoop on big Baltimore bridge crash?

I Googled to see what happened to them , one was put into service in Maine to NS for awhile , they are both in service with the Navy from what I can tell .

 

The one operating as USNS Guam is still operational. When the Army and Navy HSV development programs merged, production the JHSVs began and as the number of those increases, the ex-Hawaii Superferries will be taken out of commission; the Army had returned their leased Incat vessels earlier.

Remarkably few of the large high-speed ferries that were built around the world in the heady days of the 1990s and early 2000s are still in service today. The opening of the Chunnel had a huge impact on some that my company helped to support. Austal (AUS) and Incat-Tasmania divide up the remaining market, with the German, French, Italian and Spanish ferries no longer being built.

 

Preliminary NTSB report on Dali's power failure has been released.

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

Sal did a video on it. But he didn't comment at all on whether the ship's power distribution panel was configured normally or not, or what the risks were in using the configuration that is depicted in the video. It's no surprise that it resulted in massive knock-on failures. The idea is that the tie link (HVR) breaks upon detecting any major fault - saving half the system and maintaining all critical systems that are redundant to each side. Dali's was intended to be normally closed. They had Dali shunting all low-side power through the closed high-side tie link. It should have gone through the breakers on the same side as the generators. So they lost 100 percent of the low side power if the high side tie link opened up. Sal did mention that until five days prior, the other pair of breakers (HR2, LR2) were being used, and that only changed when a maintenance goof had people flipping switches to fix a unintended genset shutdown in port.

For the technically inclined, a bit more insight can be found in the link below under the heading Segmented Emergency Bus.
Increasing Power Redundancy https://www.ascopower.com/us/en/resources/technical-briefs/increasing-power-redundancy.jsp

Basically, this is why I hate redundancy. Sooner or later, the added complication will bite you in the butt. It's only a question of whether or not it has managed to save your butt a couple times before it does you in. A pretty good rule of thumb is that if it doubles your options for fixing stuff , it quadruples the number of ways you can screw things up. And you don't know that you screwed up until it's too late. I would very much like to read, as part of the NTSB report, the failure modes and effects analysis of the Dali's power distribution system. And I want to know what the permissible configurations are for departing port.

And there is no mention of what started it all, but it really doesn't matter. It could have been anything. The cascade was inevitable.

 

Chief MAKOi video. He mentions the differential protections system, although thinks it unlikely. It was the first thing that came to my mind with HR1 and LR1 both tripping. This loop will have a fault detection, isolation, and recovery (FDIR) system, and the transformers will have detection for internal faults that might be stand alone or integrated into the rest of the segmented loop's fault response system. These systems are expecting a limited set of "normal" starting configurations, and don't always behave well if the starting configuration is wonky. I suspect there was a serious real fault, and it may be that it couldn't have been recovered from in time under the best of circumstances. But the wonky initial conditions appear to have created knock-on failures. The non-availability of the bow thruster is one example. Might not have been enough, but it should have been there. Generator trips on 3 and 4 could also be the result of differential fault detection on the HV bus. Switching on HR1 and LR1 manually might remove the remote differential fault disconnect capability, and the fault was then detected as a differential fault on the high bus. The HVR tie breaker should open if this happens.

 

@philSweet

I do not appreciate your remark on redundancy. Meaning I don’t understand.

Suppose Dali systems were for two totally separate systems, one per engine. One system fails, the other does not and the engine on one side still functions.

In the simplest context, I get the general idea that two separate systems result in two places to fix stuff, but maybe I misunderstand.?

 

Yes, redundant systems that don't require a "recovery" action are more robust in the overall scheme of things. They are the way to go on stationary equipment. Loop systems like this with a reserve transformer are a bit different. In a perfect world, you can isolate nearly any component and still operate at full capacity - and you can isolate for maintenance. But the additional complexity means that the systems degrade differently over time. Every time they work, it breeds complacency, and the funding needed to maintain the systems adequately disappears when the these systems get passed on to second and third owners.

I should point out that the same architecture with a double fed, spare transformer, segmented loop is used in all sorts of places. Substations and heavy industry use it, so there are tens of thousands of these scattered around not including ships.

 

I just had a conversation with one of the federal investigators.

He said that by the time he (and I guess other investigators) was there to ask questions, the Maryland Pilot's Association members (there were presumably 2 MPA members on board, though one may have been an apprentice) was gone. And MPA refused to identify the pilots. Which made the investigator's job impossible to do as well as he wanted, because he couldn't ask them questions. The ship's log didn't give the pilot's full names either.

He also mentioned a prior court case. A Maryland pilot had made a mistake that led to an accident. Federal authorities had revoked his federal pilot's license, which would have also lost him his Maryland pilot's license. But he successfully challenged it in court, saying that he was operating under Maryland authority, not federal, so the federal government lacked jurisdiction. Since then, the federal government has been unable to revoke pilot's licenses in states that have their own pilots. (I think I got that right.)

He also mentioned that there were only a few crew members available at the emergency engine start location to handle reversing the engine - which involved shutting the engine down, pressurizing a cylinder from an air tank, which had limited capacity (it was recharged using a pump that ran off of the engine shaft), which might therefore have meant that by the second electrical failure, there was insufficient air pressure available to restart the engine, though he wasn't sure of that. (Interesting: some really big engines like that can actually run in reverse - without needing a transmission to do so.)

He also mentioned that it was roughly a 30 minute walk from one part to another on the ship - so they couldn't put more crew into that location quickly enough.

Rules about the minimum number of crew members available is determined by the country that registers the vessel - in this case Singapore. The U.S. basically can't regulate that. Ships often register in countries that minimize rules and costs.

He said the anchor probably couldn't steer the ship quickly enough, because it was dragging along a part of the bottom that had loose mud.

I wondered why the planned route had been so close to the bridge pylons, and why it involved turns. He said that was the path of the navigational channel, which the ship had to try to follow because it had a fairly deep draft.

I asked him if this accident was likely to change any rules. His best guess is that in the future, the Port of Baltimore will require tugs to accompany big ships past the bridge - even if that increases the cost of big ships using the port.

 

Well... that's about what I'd expect.

Most of these kinds of deals are a pile of smaller bad decisions or short cuts, and usually several safeguard bollards blown through.

Unfortunately it's cultural phenomenon, hard to identify and remove folks with sub par competency. Even harder to fix the issue as we run all over on such narrow reserves of man power.

Sadly it takes death and bridge collapse to get some follow through, and even then very little consequences.

 

I still completely fail to see anything a pilot could have done to avert this accident. I'm pleased to see the reference to the 30-minute response/travel time...just spitballing, I'd told some " why it take so long to respond with backup" wag that it would be at least a 15-minute jaunt to get to the local steering control that I'm familiar with. Admittedly, my experience is all on much smaller vessels.

 

Would that depend on whether sending away the tugs before reaching the bridge was their decision? Could they be blamed for it?

When a disaster occurs, people and companies want someone specific to blame. And every possibly responsible party is going to try to shift the blame elsewhere, including in the courts. Is trying to hide the pilots' identities until the NTSB possibly forces them out into the open the best way to protect them? If they stay hidden, others might blame them.

 

Hey, I just came up with a new way to debug AIs!

Use them to replace ship pilots and masters. Future Dalis should be autonomous vessels.

What could possibly go wrong?

 

They are going to try it out on shopping carts first. For those of us who call them buggies, it puts a whole new spin on things.
Smart Shopping Carts | AI Shopping Cart Technology | Caper https://www.caper.ai/

Sorta makes sense as an antitheft measure for stores with self checkout.

 

Very good point. Most political scandals are made into scandals because someone tried to cover up something, when if they just said, "Oh, yeah I did that. Sorry, I thought it was okay." They would likely keep their office.

Maybe that's what the Pilot's Association is hiding. They are secretly using AI and don't have an actual pilot to interview.

-Will

 

Friends of mine operate and/or crew tugs in that harbor. They said that the handling of the Dali is/was SOP, although they'd be quite happy if port authorities decided to increase tug requirements.

 

Considering the horizontal clearance at that bridge, was there enough space to have tugs on the sides of the ship?