Small fault can cause big problems

One joint stops 4 engines?

In this case it appears so. According to below info they had an HT cooling loop that was common to the four engines. That is piss poor redundancy for a ferry, even one approaching thirty years old. I’m a little surprised with four engines that there weren’t at least two separate HT loops to allow for split redundancy.

The TAIC news release said:

“The main and auxiliary engines shut down because their high temperature cooling water system had failed; this happened in part because one component, a rubber expansion joint, had ruptured and most of the cooling water drained out before the crew could stop it.”

Per Interislander Executive General Manager Walter Rushbrook said.

“As is common in ships, there is one water cooling system which is connected to all four of the main ship engines and also to the engines that run the generators for electricity.

A leak occurred in a connection that is part of the cooling system, which resulted in a loss of pressure.

Sensors in each of the engines detected the reduction in pressure and the engines automatically shut down to protect them from overheating."


another class approved success…

“Run it til it breaks” strikes again.


The Germans have an engineering term for this: “Kaputsparren.”

“Run it till it breaks” is sometimes a valid maintenance plan in the right application (obviously not this one). I’d be curious to read the final report and recommendations when published. I hope the root cause is not the failure to follow OEM guidelines for component change intervals.

"This happened because KiwiRail had not followed the manufacturer’s advice; even under KiwiRail’s own system, the REJ was two months overdue for replacement.

That statement bothers me. If I were looking at a root cause, I’d say in wasn’t the component failure or the fact that it was expired before it was even installed. It was a lack of engineered design redundancy that allowed a single component failure to cause a full black-out/loss of propulsion.

They could put all brand new flex joints straight off the assembly line into this HT CW system and if just one fails again tomorrow, whether by manufacturing flaw, installation error, someone drops a pointy screwdriver on it from above, whatever, they will end up with the same result of full ship failure. Not to mention, for a four-engine loop it seems likely there are other single point failure modes in that piping system. For a vessel rated to carry over 1,300 pax that is pretty unacceptable.


There are cruise ships out there with single point failures and proved by the operators when the dead ship full of customers gets towed to port. ( with no running water or working toilets)
Class failures again.

Please see my post in the Anatomy of a Disaster thread about emergent properties and redundancy.

Today’s managers are taught, advised, and pressured to eliminate redundancy everywhere possible. Mix in a good dose of regulatory capture and you get lurking catastrophic faults sprinkled through everything. The only way the people at the end of this flawed supply chain can cope is to be extra diligent with maintenance and if that is not permitted sh*t is bound to happen.

I saw this trend happening in 1995 in the security software field (along with the “fake it until you make it” philosophy of the Silicon Valley tech kiddies) and left the corporate world behind. Best choice I ever made.