Marine directorate retains testing firm
THE marine directorate has retained the services of Seaspeed Technology to investigate and develop an understanding of failure analysis applicable to the directional control systems of high-speed passenger craft.
The study was undertaken between June and November 1992 at a time when rapid development was taking place in the design and operation of large high-speed passenger craft with consequent revisions being proposed for the Code of Safety for Dynamically Supported Craft.
Failure analysis involves the study of the craft, its systems and its environment with the objective of defining the possible failure states that may exist and which may result in significant reductions in levels of safety to the craft and its occupants.
The study includes a detailed investigation into two modern high-speed ferries and provides an analysis framework which may be used for future investigations.
The main investigation found that a successful study may be undertaken using a combination of failure mode and effect analysis, fault tree analysis and event tree analysis techniques, while accepting that there may be other relevant methods.
The proposed study is: A functional failure analysis which defines and studies the main areas of interest and which may, for many craft, be sufficient for the administration's requirements.
A system failure analysis, which investigates systems identified in the functional analysis as requiring study at greater depth.
A complex system or component failure analysis - data from system suppliers.
Documentation summarising the study, particularly the important failure conditions.
Failure analysis sea trials - in which particular failures are induced at sea.
The report highlights the need to consider the human element in the operation of the craft and provides a technique for doing so.
This is, however, limited to the area of the system/human interface since human errors in general are considered beyond the scope of this report.
For the craft investigated it was assessed that a loss in directional control may occur once in about five years of operation, with an accident arising once in every 25 of such incidents.
The report recommends that failure analysis should form part of any future certification process, with a longer-term objective of being integrated into quantified risk assessment procedures such as safety case presentations.
It also recommends that historical marine accident data be analysed in such a way as to provide guidance for particular accident-type frequency levels and that future accident data be recorded in a way relevant to this type of safety assessment.
This is considered to be a most important requirement for the successful development and use of these probability based methods.