DUKWs? Fiat lux

I have written here about a series of misfortunes suffered by DUKWs in Liverpool and London. On 16 June 2013 I put up some photos of DUKWs in Dublin and Liverpool; I pointed to what seemed to me to be two differences between practices in the two cities:

First, before they enter the water at Grand Canal Dock, Ringsend, the DUKWs are fitted with extra buoyancy in cylinders that slide into racks along their sides. I saw the VikingSplash crew removing the cylinders from the yellow DUKW; it took only a couple of minutes, and I presume that it didn’t take much longer to put the cylinders on.

Second, the Dublin passengers are issued with buoyancy aids before they take to the water. I can’t see any buoyancy aids on the Liverpool passengers, although it’s possible that they are out of camera shot.

There are links on that page to photos, news reports and a seriously scary video of the sinking of a DUKW in Liverpool. Then, in September 2013, a DUKW went on fire on the Thames; my brief report and links here. In October 2013 the UK Marine Accident Investigation Branch issued a safety bulletin (my report here, with links) pointing to foam buoyancy as a common factor. The Liverpool DUKWs did not have enough buoyancy to keep them afloat if they started taking water and MAIB thought it would be impossible to get enough into them. On the London vessel,

… the most likely cause of fire was the action of the rotating drive shaft (or other moving parts) on the oil contaminated surfaces of the buoyancy foam blocks.

In November 2013 I noted that the wearing of lifejackets had been discussed in London and I commented on the policy of the Dublin operator, VikingSplash:

The point that strikes me is that, in both UK accidents, passengers had little time to don lifejackets and would have been trying to put them on in a confined space and under less than ideal conditions. It seems to me that Viking Splash’s policy [having passengers don lifejackets before taking to the water] is the right one.

In December 2014 the MAIB published its report into the two accidents. There’s a Guardian news report here [h/t gjb] and you can download the MAIB’s full report and annexes from this page. The London and Liverpool accidents are covered in the same report.

It’s well worth reading and pulls (as far as I could tell) no punches, even tearing strips off the Maritime and Coastguard Agency and the Vehicle and Operator Services Agency, both within the UK Department for Transport and both involved because the DUKW is an amphibian.

I took three main points from the report.

The Irish approach

First, the Irish authorities seem to have thought seriously about the safety problems. My understanding of the buoyancy requirements was mistaken: the UK DUKWs had added buoyancy (although not enough, and adding more foam caused fire) and the Irish buoyancy cylinders are not designed to keep the vessel afloat. Here’s what MAIB says:

In Ireland, APV operators have been permitted to operate vintage DUKWs without having to provide any residual buoyancy. To mitigate the consequences of serious flooding, the Irish regulator required the operators to:

  • Fit external buoyancy tubes designed to slow the sinking process and make the vehicle sink bodily [my emphasis].
  • Retract the canopy roof and open the side curtains prior to entering the water.
  • Require passengers and crew to wear PFDs while on the water.
  • Provide a fast rescue craft, rescue crew and an inflatable liferaft at the slipway.
  • Limit operations to a non-tidal area.

This approach focused on passenger survivability by reducing the risk of entrapment and drowning, rather than vehicle survivability, and introduced several of the interim measures recommended by the NTSB following the sinking of Miss Majestic.

The Irish model demonstrates that open topped APVs can be operated successfully in similar weather conditions to those experienced in the UK, and that passengers are willing to wear PFDs.

So big it up for the Irish Maritime Administration.

The speed of the sinking

Second, if you’ve seen the video of the Liverpool sinking you may have been struck by its speed. In both Liverpool and London passengers had very little time to get out and the report’s synopsis says

In both instances, the crew had little time to co-ordinate the evacuation process and the confined nature of passenger spaces made it almost impossible for them to control or assist the passengers.

And in 4.3 Common safety issues:

8.  It was extremely fortunate that all on board WQ1 and Cleopatra were able to evacuate into the water unharmed. In both cases the passengers were forced to act on instinct and exit the vehicles under their own initiative.

Any of several issues could have cut the time available and “the risk of entrapment and the likelihood of loss of life would have been considerably higher”.

So Figure 63 of a Dublin DUKW doesn’t just show the external buoyancy cylinders: the passengers are wearing buoyancy aids, there is a crewman already stationed at the stern and the side and roof canopies are open, all giving more chance of escape.

I’m not in any position to assess the overall safety of the Irish DUKW operation. What interests me here is a more general point about the evacuation of passengers from trip vessels: getting a large number of people out of a small space in a short time is not easy. And the recent problem of getting people off the Norman Atlantic didn’t make me feel any better.

Photo

Third, I am delighted that my photo of a DUKW in Liverpool was useful to the  UK Marine Accident Investigation Branch (see the report’s Figure 16) but they really should have asked for permission to use it.

 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s