Friday, 14 May 2010

FV Trident Formal Investigation – the distance from reality

Toto, I’ve a feeling we are not in Kansas anymore” (Dorothy Gale, from the film The Wizard of Oz)

With every day that passes, it becomes more and more apparent that the UK government would very much like the Trident court of inquiry to dismiss the findings of the original 1975 public investigation and conclude instead that the loss of the vessel and its seven crew was caused by some reason other than deficient stability. 
In fact, the cause for the loss that has been proposed by the inquiry’s Joint Panel of Experts (JPE), after many years of deliberation, and which the Government is vigorously promoting is that:

 “The cause of this capsize is attributed to specific sea-keeping characteristics of the vessel combined with the prevailing sea conditions at the time”

To arrive at the above conclusion, without or in spite of the available factual evidence, a few premises need to be introduced beforehand, which when you use a long enough chain of estimative processes, approximations and other abstractions of reality, and when you are not constrained by empirical verification, can be quite easy.

In order to demonstrate and produce evidence about the behaviour of the Trident in various sea conditions the Maritime Research Institute Netherlands (MARIN) was hired to construct a physical model of the vessel, which was tank tested in the weather conditions specified by the inquiry Joint Panel of Experts (JPE), as well as a Fredyn numerical model, which was tuned using the tank testing results from the physical model.

Trident’s weight and centre of gravity details, normally derived from an inclining test, used by MARIN to build their models were, however, a step further from reality since they had been obtained from sister vessel data and negotiations amongst the parties represented at the inquiry.
The weather conditions, specified by the JPE, inconsistent with several eyewitness testimonies and the findings of the original investigation [*], were derived from two weather hindcasts – i.e. other approximations of reality – and then processed for the purpose of providing the necessary parameters for the model. 
How this processing was done and how reliable its outputs were, we may never be able to fathom. All we really know is that the conclusions drawn by the inquiry experts from these hindcasts suggest that, on the day when the Trident was lost, the winds and the sea waves were much bigger than the testimony given at the time of the 1975 inquiry indicated.
What is more, the MARIN physical model was only run for a limited number of wave settings, leaving the scientists to analogise freely as to the reactions of the model to other sea conditions.

Then, of course, the error propagation comes into play and, in the end, the results obtained from this combination of successive abstractions of reality, with their accumulated errors and subjectivity, doesn’t inspire great confidence.

In short, it can be argued that testing the behaviour of a vessel whose displacement and centre of gravity at the time of her loss are not accurately known, under weather conditions the parameters for which appear to have been interpolated from extrapolations, by means of a model which incorporates a number of possibly debatable assumptions and suppositions as well as a series of further abstractions, validating this model against another model, observing it through a very limited number of tests and assessing the test results using yardsticks and norms that have not been accepted in the wider maritime community, takes us a some distance from reality and from a level of certainty than we might consider suitable to a fatal accident investigation.
[*] If we understood correctly the press reports on this subject, the victims’ families were prevented from appointing their own weather specialist. (Aberdeen Press and Journal, 02 November 2009, Trident families’ weather expert is disallowed)

No comments:

Post a Comment